T A X P O L I C Y A N D T A X A V O I D A N C E : T H E G E N E R A L A N T I - A V O I D A N C E R U L E F R O M A T A X P O L I C Y P E R S P E C T I V E by M I C H A E L B R E N D A N T A Y L O R B . A . (Hons.), The University of Alberta, 1996 L L . B . , The University of British Columbia, 1999 A THESIS S U B M I T T E D IN P A R T I A L F U L F I L M E N T O F T H E R E Q U I R E M E N T S F O R T H E D E G R E E O F M A S T E R O F L A W S in T H E F A C U L T Y O F G R A D U A T E S T U D I E S T H E U N I V E R S I T Y O F B R I T I S H C O L U M B I A Apr i l 2006 © Michael Brendan Taylor, 2006 ABSTRACT From a tax policy perspective, tax avoidance creates serious problems by undermining the key objectives and features of a good tax system. Therefore, most advanced tax systems attempt to restrain abusive tax avoidance. The General Anti-Avoidance Rule (the " G A A R " ) is the primary Canadian control on tax avoidance in the income tax context. Despite the presence of a statutory G A A R , however, Canadian tax law struggles to clearly and consistently identify abusive tax avoidance and to distinguish it from legitimate tax mitigation. In this thesis, the author proposes a theoretical framework for identifying abusive tax avoidance and for controlling it using the G A A R . The tax system is of central importance in democracies like Canada, and certain values linked to equity, fairness, and the rule of law must be respected in distinguishing between tax avoidance and tax mitigation. Drawing on common approaches to identifying tax avoidance, the thesis proposes its own theoretical approach from a tax policy perspective. Courts in other jurisdictions have adopted a variety of judicial doctrines aimed at controlling tax avoidance, such as the business purpose test, the step transaction doctrine, an economic substance analysis, and the abuse of rights doctrine. Canadian courts confronting tax avoidance have declined to adopt any such doctrines as a matter of statutory interpretation. The G A A R is intended to mandate a break from the Canadian interpretive tradition. The G A A R creates a three-step framework to identifying tax avoidance - tax benefit, avoidance transaction, and misuse or abuse. Those concepts are related to the theoretical Ill approaches and anti-avoidance doctrines reviewed in this thesis. The author suggests that the G A A R can be applied in a manner that preserves ideal tax policy outcomes in tax avoidance cases. The thesis examines real tax avoidance scenarios in which Canadian courts have resisted applying the G A A R - a circular financing sale-leaseback transaction, and surplus stripping. The thesis argues that these transactions are indeed abusive when viewed in terms of the theory proposed, and that Canadian courts have made key errors in applying the G A A R . iv TABLE OF CONTENTS Abstract i i Table of Contents iv Acknowledgements v i i I. Introduction 1 A . Understanding Tax Avoidance from a Tax Policy Perspective 1 B . Comparing Approaches to Controlling Tax Avoidance 5 C . Is the G A A R Consistent with Sound Tax Policy? 7 II. Tax Policy Problems Created by Tax Avoidance 9 A . Objectives of the Tax System 9 B . Features of a "Good" Tax System 12 i . Equity 12 i i . Neutrality 15 i i i . Simplicity 16 iv. Efficiency 17 III. The Nature of Tax Avoidance From a Tax Policy Perspective 19 A . Taxes, Democracy and Fairness 20 B . The Nature of the Tax System. 24 C . Identifying Tax Avoidance by Distinguishing Between Degrees of Behavioural Modification: Tax Avoidance, Tax Evasion and Tax Mitigation 28 i . Tax Evasion 30 V i i . Tax Mitigation 31 D . Common Approaches to Identifying Tax Avoidance 35 i . Form 35 i i . Purpose 39 i i i . Policy 43 E. A Tax Policy-Based Theory of Tax Avoidance 46 IV. Comparative Approaches to Controlling Tax Avoidance 50 A . The Importance of Statutory Interpretation in Combating Tax Avoidance 52 B . The Sham Transaction Doctrine 55 C . The Ineffective Transaction Doctrine 59 i . The Doctrine After Stubart and Continental Bank 62 i i . Rectification 65 D . The Business Purpose Test 67 E . The Step Transaction Doctrine 72 F. Economic Substance 80 i . Economic Substance in the United States 83 i i . Substance and Form in the U K and Canada 86 G . The Abuse of Rights Doctrine 91 H . Conclusions 94 V . The G A A R 98 A . Statutory Interpretation and Tax Avoidance in Canada 99 B . Outline of the G A A R - Tax Benefit, Avoidance Transaction and Misuse or Abuse 112 C. Identifying a Tax Benefit 113 D . Avoidance Transaction and "Series of Transactions" . 123 i . Objective Purposes as Compared to Subjective Motives ..125 i i . Primarily for Bona Fide Non-Tax Purposes 130 vi i i i . Series of Transactions 135 E. The Exception for Misuse, Abuse and Frustration 139 F. Conclusions 146 V I . Applying the G A A R to Ensure Appropriate Outcomes From A Tax Policy Perspective 148 A . A Circular Financing Sale-Leaseback: Canada Trustco Mortgage Company 150 i . The Facts and Judgments 150 i i . Normative Tax Consequences and Applying the G A A R 156 B . Surplus Stripping: McNichol and Evans 162 i . McNichol: The Facts and Judgment .162 i i . Evans: The Facts and Judgment 166 i i i . Normative Tax Consequences and Applying the G A A R 169 C. Conclusions 177 VII. Conclusions 180 Bibliography. 185 ACKNOWLEDGMENTS I particularly wish to acknowledge the advice and assistance with this project provided by my supervisor, Professor Claire Young of the Faculty of Law. Her insightful comments and suggestions were of great assistance in directing my thinking and refining my analysis. Her involvement has also resulted in a clearer, more incisive thesis than would otherwise be the case. I would also like to acknowledge Professor K i m Brooks of the Faculty of Law, who served as the second member of my committee. Her comments and encouragement, particularly with respect to earlier thoughts on which parts of this thesis are based, were invaluable. 1 I. INTRODUCTION A. Understanding Tax Avoidance from a Tax Policy Perspective Taxes present an interesting paradox: as individual taxpayers, virtually none of us enjoy paying taxes (at least any more than absolutely necessary); and yet, because taxes fund essential public services, collectively we need taxes to be paid. Moreover, although each of us might wish to pay less tax ourselves, we become indignant when we see other taxpayers pay less than we believe to be their "fair share." This reaction occurs because taxes, and our attitude towards them, are premised on our sense of justice. Taxes are the means by which a democratic society allocates the burden of paying for collective priorities, and the distribution of the tax burden among various members of society reflects a kind of political bargain regarding a just sharing of that cost. Concerns that other taxpayers may be avoiding or evading their rightful tax burden ultimately arise from a sense of injustice - one taxpayer asks "why should I pay so much i f others who are equally (or more) well-off pay less?" The paradox arises partly because, in a market economy, we are all economic actors and from an individual taxpayer's perspective taxes are an economic cost. Rational economic actors that human beings are, we endeavour to reduce or avoid that cost to the greatest extent possible. As a result, when a tax such as income tax is imposed, taxpayers respond by modifying their economic behaviour in ways that exploit opportunities to reduce taxes permitted under the relevant legislation (theoretically, only an economically irrational taxpayer would not at least take tax into account).1 A n income tax system presents a wide variety of such opportunities. The simplest, and most basic, option is simply to cease an 1 In a complex economy like Canada's, the tax mix includes various types of taxes, such as income tax, consumption taxes, excise taxes, property taxes, and so on. In this paper, I will focus primarily on tax avoidance in the income tax context, because income tax is the largest component of Canada's tax system. 2 activity that attracts tax, such as choosing leisure over work in order to avoid receiving taxable income (or, in a consumption tax context, choosing to stop consuming a taxed product such as tobacco or gasoline). Other tax-reduction opportunities, such as deferring income to a future time period, shifting income receipts to a different tax unit, or receiving income from tax-favoured sources such as capital gains, arise from the structure of the tax system. Further opportunities result from specific tax incentives implemented by Parliament, such as tax preferences for retirement savings, charitable donations, or particular types of capital investment. Sti l l more tax-reduction opportunities result from the fact that the legislation implementing Parliament's policy is invariably imperfect - loopholes can always be found and the rules used in ways that result in lower taxes but which might not be what Parliament anticipated.2 Finally, some taxpayers choose outright non-compliance, engaging in illegal activities that amount to fraud upon the tax authorities. In considering the range of behavioural responses to taxation, tax lawyers and academics agree that tax avoidance occurs at some point along that spectrum. "Tax avoidance" refers to engaging in otherwise legally permissible activities that result in lower taxes payable under the legislation than would have been payable had those activities not been undertaken. In addition, "tax avoidance" is often undertaken solely or primarily for the purpose of securing a tax reduction. Such a broad definition is not particularly helpful, however, because many legal activities that result in lower taxes payable are not objectionable from a tax policy perspective. Such unobjectionable tax-reduction activities 2 It appears that taxpayers have been exploiting loopholes for a very long time. Harvey cites a 1757 English case concerning a window tax, in which the Commissioners discovered that a taxpayer had fixed a short glass connector between two windows, thus converting two windows into one for purposes of reducing the tax. Brian W. Harvey, "Tax Avoidance - Illegal, Immoral or Fattening?" (1970) 21 N. Ir. Legal Q. 235 at 236. Similarly, Krishna cites a Mesopotamian king 6,000 years ago whose citizens swam the local river to avoid paying the local ferry toll tax. See Vern Krishna, Fundamentals of Canadian Income Tax, 8* ed. (Toronto: Carswell, 2004) at 993-94. 3 are generally referred to as "tax mitigation," or merely as good "tax planning." Tax mitigation activities that are not objectionable from a tax policy perspective pose no threat to the tax system and should be permitted. In fact, the tax system presumes that all taxpayers wi l l engage in some degree of legitimate tax mitigation. Unlike tax mitigation, however, tax avoidance is objectionable because it has significant adverse effects on the tax system. The second Chapter of this thesis explains how tax avoidance adversely impacts the tax system in a variety of ways. First, tax avoidance undermines the tax system's ability to achieve important societal objectives - most importantly, raising adequate revenues to fund public services and redistributing income and wealth to moderate the distribution resulting from the free market. In addition, tax avoidance undermines several key features of a good tax system emphasized by economists and tax lawyers, such as equity, neutrality, simplicity, and efficiency. Most importantly, in a democracy like Canada, the tax system implements a political bargain between citizens and their government intended to fairly allocate the burden of funding collective action undertaken through the state. One of the key elements of this bargain is that citizens are expected to fulfill their obligations in good faith and pay their fair share of the collective tax bi l l . In the long term, unfettered tax avoidance threatens to undermine public confidence in the tax system because it engenders feelings of injustice that undermine the willingness of taxpayers who do not avoid tax to continue meeting their obligations to the fullest. In a self-assessing tax system premised on widespread public compliance, this is a dangerous dynamic. 3 The phrase "tax mitigation" is widely used to describe tax-reducing transactions that are acceptable from a tax policy perspective. For instance, see Inland Revenue Commissioners v. Willoughby, [1997] S.T.C. 995 (H.L.) [I.R.C. v. Willoughby], per Lord Nolan; Challenge Corporation Ltd. v. Commissioner of Inland Revenue, [1987] A.C. 155 (P.C.) [Challenge Corporation Ltd. v. C.I.R.]; and Krishna, ibid, at 992. 4 Because tax avoidance has significant consequences for the tax system, tax laws must counter abusive tax avoidance. However, laws aimed at tax avoidance should nonetheless permit taxpayers to continue legitimate tax mitigation. The immediate challenge that presents itself, therefore, is how to distinguish between tax reductions that are permissible and tax reductions that are abusive and therefore qualify as "tax avoidance." This thesis argues that it is possible to identify tax avoidance, and to distinguish it from tax mitigation, from an explicitly tax policy-oriented perspective. Such a study asks, from a tax policy perspective, "what makes some transactions abusive while others are not?" In this vein, the third Chapter of this thesis explores various approaches to explaining, from a theoretical or tax policy perspective, what the phrase "abusive tax avoidance" means.4 The answer to this query is grounded in an appreciation that tax law is merely a legal instrument used to implement Parliament's broader fiscal and economic policies, and that tax laws, like any legislation, are imperfect. Tax laws do not always impose the tax consequences on particular economic activities that economic and fiscal policy dictate (and, presumably, that Parliament would have intended), because taxpayers may structure their affairs in myriad ways and it is impossible for Parliament to anticipate or to provide for all of them. 5 This idea is the key to formulating a theory of abusive tax avoidance - abusiveness lies in a taxpayer obtaining a tax result under the legislation that runs counter to the result that Parliament intended, or would have intended, to impose on the taxpayer's economic r circumstances. 4 Throughout this thesis, I will refer to abusive tax avoidance merely as "tax avoidance." Tax reduction activities that, from a policy perspective, are not abusive, will be referred to as "tax mitigation" or "tax planning." Furthermore, once Parliament enacts tax laws, the very wording of those laws themselves will influence taxpayers' economic choices and the structure of their transactions, often in ways not anticipated by Parliament. 5 B. Comparing Approaches to Controlling Tax Avoidance Given that tax avoidance does occur, and that it is a serious problem for the tax system, the natural next step in a study such as this one is to consider what can be done to control tax avoidance. Parliament's first response to tax avoidance resulting from the unanticipated use of particular tax rules is to enact new, more specific rules targeting precisely those tax avoidance opportunities. Unfortunately, a common result of enacting more specific rules is that motivated taxpayers, often with the aid of bright accounting and legal advisors, simply discover new avoidance opportunities created by the new rules. Such opportunities typically then lead Parliament to enact additional detailed rules aimed at the new avoidance opportunities. Ultimately, merely enacting detailed rules targeting specific transactions is an inadequate response to tax avoidance for two reasons: first, it leads to a self-propelling cycle of taxpayer avoidance and Parliamentary response, with disastrous results in terms of the complexity of the tax legislation; and, second, such an inherently reactive approach is a significant problem for a tax system that strives to be equitable. It is inequitable to regularly permit the most aggressive taxpayers, who are often the ones with the most resources, to benefit from discovering new tax avoidance opportunities but then to close those avenues before all taxpayers have an opportunity to take advantage of them. 6 A s a result, most jurisdictions have developed some form of general anti-avoidance rule. Those rules can be either statutory, such as section 245 of Canada's Income Tax Act 6 One obvious solution to the inequity resulting from reliance on specific anti-avoidance rules is to enact retroactive legislation denying the tax reduction to all taxpayers talcing advantage of a particular transaction, not merely to those who come after the legislation. Such legislation would indeed level the playing field as between taxpayers, but at a significant political cost. Furthermore, although retroactive legislation is constitutional in Canada, it is nonetheless unusual and there is a great deal of reluctance to enact it. The practice in Canada is typically that the Department of Finance will issue a press release signalling an intention to enact new rules, and the subsequent amendments will be retroactive only to the date of that press release. 6 (known as the General Anti-Avoidance Rule, or the G A A R ) , 7 or they can be extra-statutory rules applied by tax authorities and courts interpreting and applying tax legislation to the facts of a given case when a taxpayer's claim to a tax reduction is disputed. A wide variety of such doctrines exist in the laws of various Western jurisdictions, particularly the United States and the United Kingdom. However, Canadian courts have proven resistant to adopting extra-statutory anti-avoidance doctrines, except for some narrow doctrines that might appropriately not be described as anti-avoidance doctrines at all. The purpose of the fourth Chapter of this thesis is to review some of the various anti-avoidance doctrines found in other jurisdictions, and to explore the extent, if any, to which those doctrines have been applied in Canada. In addition, the fourth Chapter w i l l consider why Canadian courts have been reluctant to adopt vigorous anti-avoidance doctrines of their own. A review of various anti-avoidance doctrines is a useful prelude to an analysis of the Canadian G A A R , because the G A A R was drafted in the late 1980s as a response to the Canadian courts' refusal to adopt common law anti-avoidance doctrines that were already relatively well-developed in other jurisdictions. A s a result, the Canadian G A A R has been influenced by ideas and concepts found in other doctrines, and any review of the G A A R benefits by relating the concepts upon which it is based to their origins in other anti-avoidance doctrines. The fifth Chapter of this thesis therefore examines how the G A A R is a reaction to the Canadian judicial approach to tax avoidance, and proceeds to analyze the G A A R itself - the provision's structure, its key concepts, and the manner in which the G A A R has been interpreted by Canadian courts to date. 7 R.S.C. 1985, c. 1 (5* Supp.) [the Act]. 7 C. Is the GAAR Consistent with Sound Tax Policy? Because the G A A R is Canada's primary response to abusive tax avoidance, it is essential that the rule properly identify tax avoidance and distinguish abusive dealings from legitimate tax mitigation. In addition, the G A A R wi l l have lasting utility in Canadian tax law only if its application accords with an understanding of tax avoidance based on sound tax policy considerations. Therefore, the sixth and final Chapter of this thesis considers how the G A A R has been applied in some specific cases and, more importantly, how the G A A R should be applied in those cases in light of the definition of abusive tax avoidance proposed in the second Chapter. This thesis reaches a number of conclusions about the appropriateness of the G A A R from a tax policy perspective . One conclusion is that the G A A R properly recognizes the need to balance between controlling abusive tax avoidance and permitting legitimate tax mitigation. In addition, the G A A R as it is drafted could properly deal with tax avoidance that is abusive from a tax policy perspective, but, for various reasons, the courts have not always applied the G A A R in keeping with sound tax policy. In particular, Canadian courts have remained overly fixated on legal relations and have resisted considering the underlying economic outcomes of taxpayers' transactions when determining what tax consequences Parliament would have intended to apply to those transactions. Since an appropriate understanding of abusive tax avoidance requires an appreciation that tax law is an instrument intended to achieve other policies, and that ultimately Parliament intends tax legislation to apply tax consequences to taxpayers' underlying economic circumstances (except in very specific situations), the G A A R ought to 8 be interpreted in accordance with those policy considerations i f it is to be truly effective in controlling abusive tax avoidance in Canada. 9 II. TAX POLICY PROBLEMS CREATED BY TAX AVOIDANCE From a tax policy perspective, tax avoidance creates many problems. Most importantly, tax avoidance prevents the tax system from fully realizing its main objectives of raising revenue and redistributing income and wealth; it impairs equity in the tax system; it undermines the neutrality of the tax system by encouraging and rewarding contrived business forms; it creates complexity where simplicity is desired; it wastes administrative resources that could be devoted to more productive endeavours than the battle over tax avoidance; and it gradually undermines public confidence in the fairness of the tax system. A l l of these negative effects suggest that controlling tax avoidance is a significant priority for policy-makers. A. Objectives of the Tax System The tax system is an important policy instrument through which Parliament seeks to achieve many economic objectives. Tax avoidance is a threat to two of the tax system's primary objectives: raising revenue and redistributing income and wealth. The primary objective, or purpose, of the tax system is to raise public revenue to fund government programs such as the provision of public goods and services. This function is naturally the 8 There is a general consensus that raising revenue and redistributing income and wealth are two of the main objectives of the tax system. For instance, refer to Robin W. Boadway & Harry M. Kitchen, Canadian Tax Policy, 3rd ed. Canadian Tax Paper No. 103 (Toronto: Canadian Tax Foundation, 1999) at 51; and to Fair Taxation in a Changing World: Report of the Ontario Fair Tax Commission (Toronto: Queen's Printer for Ontario, 1993) at 35 [Report of the Ontario Fair Tax Commission]. Compare Neil Brooks, "The Changing Structure of the Canadian Tax System: Accommodating the Rich" (1993) 31 Osgoode Hall L.J. 137 at 149-51 [Brooks, "Changing Structure"], who describes the primary objective of the tax system as the reallocation of resources through the state provision of goods and services that the market does not provide efficiently, such as national defence and education. In Brooks' view, raising revenue is not an objective itself, but merely the means to achieve greater policy ends. 10 most important function of the tax system, and all other purposes (although important) are ultimately incidental because, without adequate revenues, the state would be unable to afford the expenditures necessary to finance its programs and services. 9 The main threat to the tax system from tax avoidance is therefore relatively easy to identify: to the extent that taxpayers are able to avoid paying taxes that the government expects to receive, the tax system raises less revenue than the government plans. 1 0 To the extent that portions of the tax base escape taxation, tax avoidance causes "tax leakage," resulting in a shortfall in the funds available for public goods and services. Furthermore, such a shortfall is unanticipated. Although the Act contains numerous incentives and tax expenditures that also reduce total revenue received, those expenditures are typically planned and budgeted for, meaning that Parliament compensates for them in the tax system in order to ensure sufficient revenues. Tax avoidance, by its very nature, results in revenue shortfalls that could not have been predicted and cannot be compensated for, since Parliament does not anticipate or intend any degree of tax avoidance. Tax avoidance therefore threatens the tax system's ability to raise the funds needed for public purposes and could potentially leave the government in a fiscal bind. Unfortunately, because tax avoidance is unanticipated and our tax system is a self-assessing system, the government has no way to track the scope of, and the revenue lost to, tax avoidance. Therefore, although it is safe to conclude that tax avoidance results in a loss of revenue, it is impossible to quantify the scope of this problem." 9 Report of the Ontario Fair Tax Commission, ibid, at 35. 1 0 Canada, Report of the Royal Commission on Taxation, vol. 3, Appendix A (Ottawa: Queen's Printer for Canada, 1966) at 541. Also Jeffrey Waincymer, "The Australian Tax Avoidance Experience and Responses: A Critical Review" in Graeme S. Cooper, ed. Tax Avoidance and the Rule of Law (Amsterdam: IBFD Publications BV, 1997) [Cooper, Tax Avoidance and the Rule of Law] 247 at 254; and Brian Arnold, "Reflections on the Relationship Between Statutory Interpretation and Tax Avoidance" (2001) 49:1 Can. Tax J. 1 at 26 [Arnold, "Reflections"]. 1 1 No studies appear to have been done in Canada and no annual reports like the Department of Finance's annual tax expenditure report are prepared. It appears that tax avoidance was a major factor in Canada's 1985-11 In addition to causing diminished public revenues, tax avoidance interferes with the redistribution of income and wealth that occurs through the tax system. Like most developed nations, Canada has generally accepted that a pure market distribution of income and wealth is socially undesirable, and our tax system is one of the most important policy instruments used to redistribute income from wealthier to poorer taxpayers. 1 2 The tax system contributes to the redistribution of income in two important ways: first, it provides the resources necessary for government spending that directly redistributes income; and, second, the tax system itself plays a redistributive role through progressive tax rates.1 3 Progressive taxation indirectly redistributes income from the wealthy to the poor by gradually increasing the share of the overall tax burden borne by wealthier taxpayers as their incomes rise. Tax avoidance interferes with both methods of redistributing income. First, when tax avoidance reduces government revenue, the government's capacity for direct spending on redistributive programs is reduced accordingly. Second, tax avoidance undermines the progressivity of the tax system because tax avoidance opportunities tend to be available more to high-income taxpayers than to their low-income counterparts.1 4 Tax avoidance opportunities are not equally available because salaried and wage-earning employees, whose taxes are withheld at source by their employers, have minimal opportunities for tax avoidance compared to those 1986 budget year, when the Department of Finance announced that corporate tax revenues were $ 1.2 billion lower than anticipated, largely through unexpected loss carry-forwards. Although the Act specifically permits loss carry-forwards, the government had not anticipated such aggressive use of those provisions or such a detrimental effect on overall government revenues. The government believed that many of those transactions were abusive, but at the time there was no provision in the Act to control abusive avoidance. The government identified this experience as a precipitating factor in the decision to introduce the GAAR. David A. Dodge, "A New and More Coherent Approach to Tax Avoidance" (1988) 34:1 Can. Tax J. 1 at 3. By contrast, Harvey, supra note 2 at 244-45, argues that due to the design of Western tax systems, tax avoidance can cause only a fractional loss of revenue and is no more than a marginal consideration. 1 2 Boadway & Kitchen, supra note 8 at 51. Also Brooks, "Changing Structure", supra note 8 at 141. 13 Report of the Ontario Fair Tax Commission, supra note 8 at 37. 14 Report of the Royal Commission on Taxation, supra note 8 at 543. Also Tax Law Review Committee, Tax Avoidance (London: The Institute for Fiscal Studies, 1997) at 4; and Arnold, "Reflections", supra note 10 at 26. 12 taxpayers earning income from business, property and capital. The vast majority of Canadians, especially lower-income Canadians, earn salary and wage income and very little income from business, property or capital, and therefore they have very few opportunities to engage in tax avoidance. To the extent that high-income taxpayers avoid more taxes than lower-income taxpayers, they reduce their relative share of the overall tax burden, shifting that burden to lower-income taxpayers and undermining the overall progressivity of the tax system. 1 5 B. Features of a "Good" Tax System In addition to interfering with the key objectives of raising revenue and redistributing income, tax avoidance undermines a number of the features that tax experts and economists agree contribute to a good tax system, including equity, neutrality, simplicity and efficiency. A "good" tax system strives to satisfy all of these criteria, and tax avoidance negatively affects each of them. i . Equity Equity refers to the need to fairly distribute the tax burden among taxpayers according to their circumstances. 1 6 Equity in taxation is generally regarded as having two 1 5 Although he does not cite any specific studies, Waincymer, supra note 10 at 257, asserts that "an analysis of tax incidence over the years tends to show that regardless of the basic political choice of progressivity in most tax systems, wealthier individuals simply do not tend to pay the expected high marginal rates." This problem was also identified by the Canadian Royal Commission on Taxation, supra note 10 at 542, and by the UK Tax Law Review Committee, supra note 14 at 4. 1 6 Boadway & Kitchen, supra note 8 at 53. !3 aspects: horizontal equity and vertical equity. Horizontal equity is the principle that taxpayers in similar economic circumstances should bear an identical tax burden. 1 7 Horizontal equity is a fundamental characteristic of the tax system because the idea that taxpayers with equal capacities to pay should pay tax equally expresses the most basic notion of fairness. In fact, Brooks goes as far as to assert that this form of equity is the single most important criterion with which to evaluate the tax system. 1 8 Vertical equity, on the other hand, presumes that the tax system ought to impose different burdens on taxpayers with different capacities to pay. 1 9 Vertical equity suggests that taxpayers with a greater capacity to pay tax should bear a greater incidence of taxation. Vertical equity underlies the progressive tax system, which places a greater relative tax burden on higher-income taxpayers. Tax avoidance has pernicious effects on both horizontal and vertical equity in the tax system. Tax avoidance offends horizontal equity because it permits taxpayers to avoid the same tax treatment as other taxpayers in similar economic circumstances, who would otherwise be considered to have an equal capacity to pay. A n income tax system measures taxpayers' capacity to pay tax in terms of income. Because opportunities for tax avoidance are not equally available to all taxpayers, those earning certain kinds of income are able to avoid taxes in ways that taxpayers earning the same amount of income in other forms are not. The result is a failure of horizontal equity, occurring when taxpayers with identical measures of economic capacity to pay bear unequal tax burdens. Thus, because opportunities for tax 17 Ibid, at 53. Also Report of the Ontario Fair Tax Commission, supra note 8 at 53. 1 8 Brooks, "Changing Structure", supra note 8 at 181. 19 Report of the Ontario Fair Tax Commission, supra note 8 at 53. 14 avoidance are not shared equally by all taxpayers, "tax avoidance is essentially 20 inequitable." Tax avoidance also offends vertical equity because sophisticated tax avoidance opportunities are disproportionately available to a minority of high-income and corporate taxpayers who receive the types of income that permit them to take advantage of those opportunities and who can afford good legal and tax advice. 2 1 Those taxpayers are the ones whom, vertical equity suggests, it is fair to charge with a greater tax burden, as occurs under a progressive tax system. However, when those taxpayers engage in tax avoidance that is unavailable to lower-income taxpayers, the relative tax incidence shifts toward lower-income taxpayers and the overall progressivity of the tax system declines. Thus, tax avoidance undermines vertical equity. Moreover, since the government's fiscal needs do not necessarily change when the tax burden borne by higher-income taxpayers decreases, Parliament must raise the rates paid by the remaining taxpayers in order to make up the revenue shortfall. The resulting rates are higher than would be necessary i f the avoidance had not occurred and, from a vertical equity perspective, are unfair to the remaining majority of taxpayers. 2 2 2 0 Malcolm Gammie, "Tax Avoidance and the Rule of Law: A Perspective from the United Kingdom" in Cooper, Tax Avoidance and the Rule of Law, supra note 10, 181 at 206. 2 1 Dodge, supra note 11 at 3. 22 Ibid, at 3. Also Michael Brooks & John Head, "Tax Avoidance: In Economics, Law and Public Choice" in Cooper, Tax Avoidance and the Rule of Law, supra note 10, 53 at 71; and Report of the Royal Commission on Taxation, supra note 10 at 542. 15 i i . Neutrality A neutral tax system would interfere as little as possible with taxpayers' business decisions and economic activities by treating alternate courses of action as identically as possible. 2 3 This principle is based on the assumption that the market economy is generally efficient and, in the absence of taxes and other market distortions, wi l l produce products at the lowest cost and allocate them in accordance with the highest economic uti l i ty. 2 4 A neutral tax system, therefore, would not unduly influence taxpayers to change otherwise efficient economic behaviour. Neutrality does not go so far as to require that the tax system should not encourage or prefer any particular economic activities, however, since Parliament intentionally uses tax incentives for many such purposes. Neutrality merely requires that the tax system should induce taxpayers to change their business decisions only when that is the government's explicit purpose. Dodge describes this principle in the following terms: "business opportunities, rather than tax planning, should be the driving force behind business and investment decisions." 2 5 Tax avoidance undermines the neutrality of the tax system because, to the extent that the tax system induces taxpayers to structure their affairs through contrived, artificial legal forms that would not be used except for tax reasons, it unduly influences their economic choices. As any survey of tax avoidance cases reveals, tax avoidance transactions often involve unnatural, contrived business forms chosen for no bona fide purpose other than to secure a tax advantage.2 6 Tax-avoiding taxpayers also engage in transactions that would never have been undertaken if not for tax motivations, even at the 23 Report of the Ontario Fair Tax Commission, supra note 8 at 64. 2 4 Boadway & Kitchen, supra note 8 at 73. 2 5 Dodge, supra note 11 at 3. 2 6 For examples, refer to Chapter 6 for discussion of a circular financing sale-leaseback scheme, a surplus stripping scheme using a transitory corporation, and a surplus stripping scheme using a conduit partnership. 16 expense of sizeable transaction costs. Tax avoidance therefore results in "a reallocation of resources to tax-preferred activities and [their] costs," which is another failure of neutrality in the tax system. hi . Simplicity Simplicity is another desirable feature of a good tax system. A s Brooks observes, "everyone agrees that a tax system should be as simple as possible." 2 8 Simplicity is important in a tax statute because it facilitates administration and compliance, it contributes to understanding and achieving other objectives of the tax system, and it helps meet the requirements of equity, neutrality, and so on. Tax avoidance is a clear threat to legislative simplicity. A s tax avoidance transactions become more complex, the legislative response required to combat them accordingly becomes more complex. However, ultimately it is impossible to anticipate all possible moves by taxpayers. As Wheatcroft comments, "even the most accurate draftsman of a law wi l l not always be able to find precise language to convey his meaning and the wisest legislator cannot foresee every possible situation that may arise." Invariably, moves by Parliament to bar specific tax avoidance schemes open new loopholes to be exploited, which in turn spawns even more sophisticated avoidance schemes that themselves require a more detailed regulatory response. The result is a cycle or "spiral" of ever-increasing legislative complexity that does not actually stop tax avoidance - as the 2 7 Waincymer, supra note 10 at 256. 2 8 Brooks, "Changing Structure", supra note 8 at 191. 29 Report of the Ontario Fair Tax Commission, supra note 8 at 65-66. 3 0 G.S.A. Wheatcroft, "The Interpretation of Taxation Laws with Special Reference to Form and Substance" (1965) 7 Cahiers de droit international fiscal, cited in Graeme S. Cooper, "Conflicts, Challenges and Choices -The Rule of Law and Anti-Avoidance Rules" in Cooper, Tax Avoidance and the Rule of Law, supra note 10, 1 at 18. Also Arnold, "Reflections", supra note 10 at 27. 17 Royal Commission on Taxation phrased it, "particularization breeds avoidance," and this approach "wins the battles but loses the war." 3 1 In the long run, tax avoidance brings Parliament into a running battle in which simplicity and accessibility of tax legislation are the casualties. iv. Efficiency Efficiency is another desirable feature of the tax system. Generally, an efficient tax system wastes a minimum amount of resources (both state resources and those of private taxpayers) on economically unproductive battles between the government and taxpayers over the administration of tax laws. Tax avoidance reduces the efficiency of the tax system because legal battles over avoidance activities consume a great deal of resources in terms of money, time and effort spent by many of Canada's brightest lawyers, accountants and administrators, which resources could be more productively spent on other pursuits. Because this activity is not economically productive, this expenditure of resources may be described as a "deadweight loss" on the economy. Unfortunately, however, much of the complexity in the tax system does not relate to outright tax avoidance, but rather merely reflects the scope and complexity of Canadians' economic activities. Therefore, even if tax 31 Report of the Royal Commission on Taxation, supra note 10 at 554-55. Also Tax Law Review Committee, supra note 14 at 4-5. The Supreme Court of Canada recognized this problem in Stubart Investments Ltd. v. R., 84 D.T.C. 6305 at 6324. Krishna, supra note 2 at 1005-06, also identifies this problematic cycle of loopholes and complexity. 32 Report of the Royal Commission on Taxation, supra note 10 at 542. 3 3 Brooks, "Changing Structure", supra note 8 at 96. On the other hand, Harvey, supra note 2 at 245, challenges this view, arguing that the time spent by lawyers and accountants would not necessarily be more valuably spent on other activities. Furthermore, Harvey argues that to a certain extent the presence of lawyers and accountants is necessary to the smooth functioning of the self-assessment system because taxpayers need their advice (and the government needs them to receive that advice) to ensure that they properly comply with their obligations to file returns, report all income, and so on. 18 avoidance disputes disappeared entirely, tax law would remain enormously complex and even attempts to comply in good faith would still require expensive professional advice. Finally, tax avoidance is a threat to public confidence in the integrity of the tax system. As a self-assessing system, our tax system depends on taxpayers' widespread faith in the integrity of the tax system in order to function smoothly. Taxpayers' perceptions of the integrity of the tax system are closely tied to their perceptions of justice and fairness. When taxpayers perceive that others are unfairly avoiding their share of the national tax b i l l , shifting their tax burden to others, taxpayers quickly become disillusioned themselves, leading to what the Royal Commission on Taxation referred to as a "deterioration in tax morality" (an increasing unwillingness to comply in good faith). 3 4 The Commission even suggested that taxpayers who believe the system is unjust but who do not themselves have the opportunity to avoid tax through legal means wi l l be more likely to engage in tax evasion, although no evidence has been gathered to determine whether such a problem does exist. However, it is clear that, whether or not a trend toward increased tax evasion is underway, preserving the integrity and credibility of our tax system is essential because it is based on self-assessment. Therefore, tax avoidance is a problem that must be dealt with in order to protect the long-term well-being of our tax system. 3 6 Report of the Royal Commission on Taxation, supra note 10 at 542. 35 Ibid, at 542. Although the Commission's suggestion may seem alarmist, it is reasonable to suggest that a widespread perception that tax avoidance opportunities are unequally available will undermine taxpayers' sense that the system is fair and that all citizens pay their fair share, which would surely have negative effects on taxpayer compliance. Refer to Kenneth W. Gideon and Ruth E. Kent, "Mrs. Gregory's Northern Tour: Canadian Proposals to Adopt the Business Purpose Test and the Step Transaction Doctrine," in Report of the Proceedings of the Thirty-Ninth Tax Conference, 1987 Conference Report (Toronto: Canadian Tax Foundation, 1988), 7:1 at 7:13. 3 6 Brooks & Head, supra note 22 at 71. 19 III. THE NATURE OF TAX AVOIDANCE FROM A TAX POLICY PERSPECTIVE This Chapter formulates a definition of tax avoidance from a tax policy perspective. The goal of this exercise is to explain why some transactions that result in tax reductions are abusive, and therefore must be controlled, while others are not. In order to arrive at this understanding, a number of factors must be considered. First, the nature of tax law as a policy instrument cannot be overstressed. Tax lawyers tend to forget that tax law is merely an expression of Parliament's underlying economic and fiscal policies. Parliament elects to apply tax consequences to particular economic circumstances for a reason. Therefore, in determining whether it is appropriate to allow taxpayers to benefit from transactions that result in tax reductions, we must consider whether their transactions have created the underlying economic circumstances to which Parliament intended to apply the sought-after tax consequences. In addition, we must acknowledge that, in a complex modern tax system, tax mitigation is acceptable and proper because there are many circumstances in which Parliament intentionally grants tax reductions. In addition to situations in which taxpayers' underlying economic circumstances are those that Parliament intends to result in tax reductions, there are numerous situations in which Parliament uses tax incentives to encourage them to engage in particular activities, or in which Parliament expressly suspends the normal operation of tax rules for other policy reasons. 3 7 Therefore, any formulation of One example is the rules in s. 70 of the Act permitting testamentary roll-overs of qualifying farm property to spouses and minor children, which suspend the operation of the normal capital gains rules in pursuit of a policy of encouraging family farm ownership. Another is the tax-free compounding of retirement savings within RRSPs and registered pension plans, which is intended to encourage private retirement saving. 20 abusive tax avoidance must provide for legitimate tax mitigation and ensure that it is not inappropriately considered abusive. Many attempts have been made to define abusive tax avoidance. Typically, the various definitions advanced are premised on one of three main approaches to identifying tax avoidance - those that focus on the legal form of transactions, those that focus on the purpose of transactions, and those that focus on the purpose of the legislative provisions that apply to a particular transaction. This Chapter wi l l review the theory underlying these various approaches, identifying the various strengths and weaknesses of each. This Chapter then concludes with my own suggestions for a sounder definition of tax avoidance in light of a number of relevant tax policy considerations, such as the nature of tax law as a policy instrument and the need to preserve opportunities for legitimate tax mitigation. A. Taxes, Democracy and Fairness In a democracy such as Canada, important decisions about political and social questions such as which goods and services to provide collectively, whether and how income and wealth earned in the free market should be redistributed from well-off to less well-off citizens, and whether certain economic activities should be encouraged or discouraged, are traditionally made through the political process. These political decisions are based on fundamental ideas about the kind of society we want, which reflect our political and social values and our conceptions of justice. In a democracy, the tax system is particularly important because it is a central policy instrument used to implement those decisions, and to raise the public funds needed to do so. Many important elements of the tax system, such as 38 Report of the Ontario Fair Tax Commission, supra note 8 at 32. 21 the incidence of taxation (the distribution of the tax burden among the members of society), themselves involve political decisions that in a democracy are based on a society's collective beliefs about fairness and justice. 3 9 "Fairness" is a particularly important concept in tax policy because, although taxes are not the only possible method of raising public funds, they are, as Brooks observes, preferable to other forms of raising revenue because they are premised on notions of fairness. 4 0 Taxes are intended to be a fair way of raising public revenues from private citizens. One aspect of fairness in tax policy relates to the incidence of the tax system. A "fair" tax incidence usually involves balancing two perspectives on justice: first, taxing based on taxpayers' respective capacity to pay tax; and second, taxing based on a connection between taxpayers' tax paid and their receipt of or benefit from public services. 4 1 The prevailing view in Canada is that justice requires spreading the tax burden among taxpayers based primarily on their ability to pay - that is, that those with a greater capacity to pay should bear a larger relative share of the overall tax burden. However, the second perspective on justice suggests that those citizens who benefit from public services should contribute to their cost. 4 2 Both principles may be found in the Canadian tax system to varying degrees - the ability to pay principle through the progressive income tax system and the benefit principle through various user fees and premiums - but Canada has adopted the ability to pay principle as the primary factor in determining a "fair" incidence of income taxation. 4 3 Regardless how a society actually allocates its tax burden, however, public 3 9 Cooper, supra note 30 at 14. 4 0 Brooks, "Changing Structure", supra note 8 at 181. 4 1 For a good discussion of these two elements of tax fairness, see the Report of the Ontario Fair Tax Commission, supra note 8 at 45-57. 4 2 Correspondingly, this view suggests that citizens who do not benefit from public services should bear little, if any, of their cost. 43 Report of the Ontario Fair Tax Commission, supra note 8 at 45-57. confidence in their tax system depends on that allocation being fair. As I argued in Chapter 2, tax avoidance undermines the fairness of the tax system in important ways. Another aspect of tax fairness relates to the fact that taxes are compulsory contributions to the fisc. In a market-oriented economy like Canada's, there exists a fundamental tension between private economic liberty and the state's need to raise public revenue through taxation. A n economically liberal society presumes that individuals have an absolute right to enjoy their private property and holds that state interference with that property should be minimized. This ideological perspective underlies several key legal principles, which together are commonly expressed as the "rule of law." One such principle is that there is no common law of taxation - the only taxes that may be imposed must be approved by Parliament. 4 4 A second principle is that governments may act only in accordance with the law. 4 5 In the tax context, the "rule of law" has traditionally been taken to mean that tax law should be clear and unambiguous, and that any doubt in the application of tax law ought to favour individual freedom from taxation. Clear and unambiguous tax laws, this argument proceeds, enhance the rule of law because the state should not be permitted to deprive private individuals of their property except when the law clearly authorizes it to do so, and because, i f tax laws are clear, taxpayers may rely on those laws to anticipate their tax liability without the risk of the state changing the rules in mid-stream. 4 6 4 4 Harvey traces this rule back to the Magna Carta and the English Bill of Rights of 1688. Harvey, supra note 2 at 235. 4 5 Cooper, supra note 30, discusses this principle at 15. Cooper also explains that, in discussions of anti-avoidance rules, a third element of the rule of law is often invoked - that the purported law must be "law-like," meaning that it must be adequately specific and its operation must be predictable. 4 6 Edward Troup, "Unacceptable Discretion: Countering Tax Avoidance and Preserving the Rights of the Individual" (1992) 13:4 Fiscal Studies 128 at 129. The principle that taxpayers should be able to rely on clear rules to predict tax outcomes without the risk that the state will later arbitrarily change the tax outcome for policy reasons that are not spelled out in the legislation is another aspect of a fair tax system. Philipps contends that these "rule of law" considerations are informed by, and are even intimately related to, a classical liberal ideology favouring private property over state action. 4 7 Construing tax provisions strictly in accordance with the rule of law, she argues, "[plays] out a classical liberal vision of social justice." Philipps argues that such a vision is premised on a "substantive value commitment to a social order characterized by minimal government interference with private power," 4 9 and that taxes are seen principally as "an impingement on liberty." 5 0 B y emphasizing the importance of the rule of law, therefore, tax law has traditionally privileged private property over the state's need for revenue, if only through the residual presumption in favour of freedom from taxation except in clear, unambiguous cases. 5 1 Ideologically-based rule of law concerns have led to the development in Canadian tax law of a presumption that taxpayers have the right to arrange their affairs as they wish in order to ensure that their taxes wi l l be as low as possible. 5 2 This presumption is most clearly expressed in Lord Tomlin's famous dictum from I.R. C. v. Duke of Westminster that "every man is entitled i f he can to order his affairs so that the tax attaching under the appropriate 4 7 Lisa Philipps, "The Supreme Court of Canada's Tax Jurisprudence: What's Wrong With The Rule of Law" (2000) 79 Can. Bar Rev. 120. 4SIbid. at 127-28. 49 Ibid, at 125-26. 50 Ibid, at 128. 5 1 Philipps advocates a competing view of the rule of law, which she suggests does not necessarily privilege private property over collective needs. Ibid, at 125-26. Although I agree with Philipps' analysis of the ideological foundations of the "rule of law," I do not intend to endorse her alternative view. I proceed on the assumption that, regardless whether the traditional rule of law considerations are appropriate or inadequate, the values represented by those concerns are important to Canadian society in general and therefore must be accounted for in constructing a theory of tax avoidance. 5 2 A number of authors observe that this presumption is almost universal in Western legal systems. Cooper, supra note 30 at 117. Also Frans Vanistendael, who surveys a number of Western European jurisdictions in "Judicial Interpretation and the Role of Anti-Abuse Provisions in Tax Law" in Cooper, Tax Avoidance and the Rule of Law, supra note 10, 131 at 132; and V. Uckmar, "Tax Avoidance/Tax Evasion" (1983) 68a Cahiers de droit fiscal international 15 at 26. Acts is less than it would otherwise be . ' O J Furthermore, tax law has traditionally rejected arguments that fairness, justice or morality militate in favour of restraining the scope of taxpayers' tax planning. For instance, in Newman, the American Justice Learned Hand dismissed morality in the following words: ...there is nothing sinister in so arranging one's affairs as to keep taxes as low as possible. Everybody does so, rich or poor, and all do right, for nobody owes any public duty to pay more than the law demands: taxes are enforced exactions, not voluntary contributions. To demand more in the name of morals is mere cant. 5 4 Although tax law has matured since the era of Duke of Westminster and Newman, and the public policy ramifications of tax avoidance are better appreciated today, the fact remains that ours is primarily a market-based society. A sound theory of tax avoidance should therefore acknowledge and incorporate, not ignore, this context and the need to balance the competing values of private property and collective demands. B. The Nature of the Tax System A theoretical understanding of tax avoidance begins with the premise that tax law is merely an instrument of broader fiscal and economic policies. 5 5 The primary purpose of tax law, naturally, is to raise the revenue needed to fund public services. In designing a tax system to raise that revenue, however, Parliament makes innumerable other policy choices about how economic activities, and in particular the economic outcomes of taxpayers' transactions, should be taxed. Those choices are then implemented through the legislation. Some policy choices relate to how the burden of funding public services should be shared 5 3 [ 1936] A.C. 1 (H.L.) at 19 [Duke of Westminster]. 54 Commissioner of Internal Revenue v. Newman, 159 F.2d 848, 850-51 (2d Cir. 1947) [Newman]. 5 5 Brooks, "Changing Structure", supra note 8 at 140-41. 25 among society's various constituents. These choices include choosing between income and consumption tax bases; and between progressive, proportional or flat rate structures. Other policy choices relate to the implementation of fiscal and economic policies through the tax system, such as by giving differential treatment to different sources of income (for instance, encouraging investment in equities by favouring capital gains over regular investment income); or by providing tax preferences for certain activities (such as offering tax credits as incentives for retirement savings, scientific research and experimental development, or f i lm-making, or accelerated capital cost allowance to encourage the acquisition of environmentally-friendly technologies). Policy choices of this second kind generally involve Parliament's using the tax system as a means to regulate economic activities in the marketplace. The common feature of this wide range of Parliamentary policy choices is that they are implemented through the tax system, in which case tax law is used merely as a means to achieve greater policy ends. Recognizing the nature of tax law as a tool for achieving broader policy objectives is critical to a theory of tax avoidance because tax avoidance typically results when the tax treatment imposed by the legislation on the economic outcome of a taxpayer's transaction is not the treatment that, from a fiscal or economic policy perspective, would have been intended. A n example of this scenario occurred in Hill v. The Queen,56 which concerned the deduction of interest. The Act provides that simple interest (interest on borrowed principal) is deductible either when paid or when legally payable, but that compound interest (interest on interest) is deductible only when paid. 5 7 B y a complicated series of transactions, H i l l 5 6 2002D.T.C. 1749 (T.C.C.) [Hill]. 5 7 Compare paragraphs 20(1 )(c), which permits a taxpayer to deduct simple interest "paid in the year or payable in respect of the year," and 20( 1 )(d), which only permits a taxpayer to deduct compound interest "paid in the year." 26 managed to transform accrued interest into principal with the result that, under the relevant provisions, interest that was really interest on previously accrued interest rather than interest on borrowed principal was deductible when payable, not merely when paid. H i l l achieved this result even though the relevant legislative provisions reflect a clear policy choice to treat simple and compound interest differently. H i l l therefore managed to obtain a favourable tax result that did not accord with Parliament's underlying fiscal policy of treating simple and compound interest differently. Such a situation is properly described as tax avoidance. Transactions like this are abusive because they frustrate the ultimate intent of the legislation by achieving a tax result that conflicts with the fiscal and economic policy underlying the legislative provisions used by the taxpayer. 5 8 Many instances of tax avoidance result when taxpayers engage in transactions that Parliament did not anticipate when drafting legislation to implement its fiscal and economic policies. Obviously, i f Parliament did not anticipate a particular use of a provision, it could not have provided for that use. This does not mean, however, that it is impossible to determine how the economic outcome of a transaction should be taxed. Regarding tax law as a tool for achieving broader policy objectives implies that it is possible to identify a normative ideal tax treatment that Parliament would have intended had Parliament anticipated the transaction. 5 9 That normative ideal tax treatment should be based primarily on the economic outcome of the transaction and how the legislation, as structured, intends to tax that economic outcome. Thus, with respect to Hill, one could argue that since the 5 8 In Hill, the Tax Court held that the GAAR did not apply because the Crown had not identified a clear and unambiguous policy underlying the legislation, in light of which the transaction could be said to result in a misuse or abuse of the legislation. For an interesting discussion of the Hill case in light of the GAAR, refer to Brian Arnold, "The Long, Slow Steady Demise of the General Anti-Avoidance Rule" (2004) 52:2 Can. Tax. J. 488 at 500-01 [Arnold, "Long, Slow Steady Demise"]. 5 9 It is of course arguable that, given the complexity of tax legislation and the arcane language used in detailed tax provisions, "Parliamentary intent" can not reliably be identified in tax law. See the discussion under the heading "Common Approaches to Identifying Tax Avoidance - Policy." legislation clearly intends compound interest to be deductible only when paid, the taxpayer should not have succeeded even though Parliament did not specifically provide for transactions that legally convert accrued interest into principal. With this approach in mind, tax avoidance can be described as an "action taken to reduce or defer tax liabilities in a way that Parliament plainly did not intend, or could not possibly have intended had the matter been put to i t . " 6 0 Unfortunately, the complexity of tax law and the fact that Parliament uses tax law in pursuit of many often conflicting policy objectives are obstacles to identifying the normative tax treatment of a given economic outcome. It can be very difficult to identify a clear intent behind many provisions of the Act. In particular, as Troup argues, in highly complex situations the legislation is a "complex web" of taxing and reliving provisions and it is very difficult to determine what the ultimate tax outcome should be. 6 1 This concern was also raised by the Supreme Court of Canada in Stubart Investments Ltd. v. R.,62 in which the Court rejected a general business purpose requirement in part due to concerns about the confusing mix of fiscal and economic policies present in the Act.63 Despite these concerns, however, most instances of tax avoidance involve the application of discrete rules deliberately triggered by particular transactions for the purpose of obtaining particular tax results, rather than situations where the overall tax liability of a taxpayer is in issue. Generally, in the context of a particular transaction, it should be possible to identify the underlying economic and fiscal policies involved, particularly where the taxpayer uses specific provisions, and Tax Law Review Committee, supra note 14 at 3. 6 1 Troup, supra note 46 at 136. 62 Supra note 31. 63 Ibid, at 6322, para. 55. 28 even i f those policies are conflicting. Therefore, on balance, a normative ideal tax treatment can probably be identified in the vast majority of tax avoidance situations. C. Identifying Tax Avoidance by Distinguishing Between Degrees of Behavioural Modification: Tax Avoidance, Tax Evasion and Tax Mitigation A useful way to begin a theoretical formulation of tax avoidance is by distinguishing tax avoidance from other behavioural responses to taxation, such as tax evasion and tax mitigation. A s explained at the outset of this thesis, taxation induces people to change their economic behaviour 6 4 because taxpayers are reluctant to pay more tax than they have to. 6 5 In economic policy terms, many responses to taxation are merely harmless behavioural modifications, such as the substitution of lower-taxed activities for higher-taxed ones. 6 6 For instance, such economic substitutions occur when individuals choose leisure over additional work in order to avoid additional taxes on the increased income, or when individuals elect to stop smoking in order to avoid tobacco taxes. From a tax policy perspective, this type of behavioural response is generally regarded as unproblematic, since i f taxpayers truly forego the benefit of income (or of consuming tobacco), then they have incurred an economic sacrifice and should be taxed accordingly. Other taxpayers, however, wi l l respond in more extreme ways that permit them to continue to enjoy the benefits of income or of consumption, but which result in a tax reduction under the literal terms of the applicable legislation. For instance, labourers electing to work overtime might arrange for their wages to be paid to a lower-taxed spouse who later gifts the cash to them, or smokers may choose to smoke cigarettes acquired and supplied tax-free by someone else. Sti l l other taxpayers may 6 4 J.A. Kay, "The Economics of Tax Avoidance" [1979], no. 6 British Tax Review 354 at 355. 6 5 Tax Law Review Committee, supra note 14 at 2. 6 6 Brooks & Head, supra note 22 at 54. 29 engage in outright non-compliance, for instance by working under the table and not reporting their income, or by purchasing black market tobacco. 6 7 The various behavioural responses to taxation may conveniently be located on a spectrum of conduct, with the least aggressive responses at one end and the most aggressive at the other. A t the least aggressive extreme are taxpayers who, for whatever reason, whether disinterest or lack of opportunity, take no steps to reduce their tax liability. These are the taxpayers whom Baker refers to as "tax-disinterested." At the other extreme are taxpayers who engage in illegal non-compliance, or tax evasion. In between these two extremes are the murky concepts of tax mitigation and tax avoidance, tax mitigation being those activities that are acceptable and tax avoidance being those that are abusive. The challenge is in identifying the boundaries between tax avoidance and tax evasion at one end, and between tax avoidance and tax mitigation at the other. 6 9 The distinction between avoidance and mitigation has been the subject of fierce debate over the years, in large part because whether taxpayers should be taxed even though the legislation does not literally cover them is, ultimately, a political question. Therefore, it is often the case that "one person's tax avoidance may be another person's prudent and sensible tax planning." 7 0 As Kay describes it, "the man who reacts to the [tax] system by drinking coca-cola instead of whisky is not engaging in tax avoidance, though he is reducing the amount of tax he pays; the man who reacts by staying at the office to drink the firm's whisky instead of his own is avoiding tax. (The man who responds by taking the firm's whisky home is engaging in tax evasion.)" Kay, supra note 64 at 355. 6 8 Philip Baker, "Tax Avoidance, Tax Mitigation and Tax Evasion," online: Gray's Inn Tax Chambers, London . 69 Ibid. 7 0 Tax Law Review Committee, supra note 14 at 3. Also Waincymer, supra note 10 at 254. 30 i . Tax Evasion The distinction between tax avoidance and tax evasion is generally regarded as relatively clear and easily articulated. Tax evasion refers to illegal, fraudulent actions undertaken to mislead the tax authorities and to prevent them from effectively administering the legislation. The key distinction between tax evasion and tax avoidance is that tax evasion 71 72 involves illegal activities and tax avoidance does not. Tax evasion is, at heart, fraud. It involves a deliberate misrepresentation to, or misleading of, the tax authorities, by such means as destroying or fabricating records, keeping parallel accounts, failing to report income, or smuggling. 7 3 Most countries treat intentional tax evasion as a criminal offence. 7 4 Tax avoidance, by contrast, involves the use of legal activities to bring about a resulting reduction in taxes payable under the legislation. From a theoretical perspective, the distinction between tax avoidance and tax evasion may be neatly summarized in terms of the existence of a tax liability: tax evasion occurs when taxpayers avoid the payment of tax liabilities existing under the legislation, while tax avoidance occurs when taxpayers' 75 transactions avoid the creation of tax liabilities in the first place. 71Report of the Royal Commission on Taxation, supra note 10 at 538. 7 2 Montgomery B. Angell, "Tax Evasion and Tax Avoidance" (1938) Colum. L. Rev. 80 at 85. Some authors even advocate changing the terminology from "tax evasion" to "tax fraud" in order to highlight the fraudulent nature of tax evasion and to accord with the French term for tax evasion, "fraud fiscale." See Adrian J. Sawyer, "Blurring the Distinction Between Avoidance and Evasion - The Abusive Tax Position" [1996], no. 5 British Tax Review 483 at 495-96, and Baker, supra note 68. 7 3 Vanistendael, supra note 52 at 131. 74 Ibid, at 131. See also Brian Arnold & James Wilson, "The General Anti-Avoidance Rule - Part I" (1988) 36:4 Can. Tax. J. 829 at 872-73 [Arnold & Wilson, "Part I"]. 7 5 Alan Gunn, "Tax Avoidance" (1978) 76:5 Mich. L. Rev. 733 at 734. Also Uckmar, supra note 52 at 20. 31 i i . Tax Mitigation The distinction between tax avoidance and tax mitigation, on the other hand, is not so clear. There are many legal ways to arrange one's affairs that w i l l reduce tax, and determining which of those are acceptable tax planning and which are abusive tax avoidance is indeed challenging. Drawing the line between the permissible and the abusive involves balancing a number of factors. The first factor is the presumption, grounded in a general belief in the importance of private property in an economically liberal society, that taxpayers should be entitled to arrange their affairs to minimize their tax burden. The second factor is that the rule of law suggests that taxpayers are entitled to rely prospectively on the terms of the legislation, and should not risk the state subsequently invoking policy concerns not reflected in the legislation. The third factor is that tax law is filled with incentive provisions by which Parliament specifically authorizes, and sometimes even encourages, taxpayers to take advantage of opportunities for tax reduction. Tax incentive provisions, such as the deduction for R R S P contributions and the small business deduction, 7 6 are known as "tax expenditures," which term refers to the use of the tax system to indirectly deliver government 77 subsidies that would otherwise be delivered through direct grants from the government. Balanced against these factors is the public interest in protecting the integrity of the tax system, from leakage of the tax base and from being undermined by abusive tax planning that uses the legislation in unintended and unanticipated ways that frustrate Parliament's aims. 7 6 Under s. 125 of the Act, Canadian-controlled private corporations enjoy a reduced rate of tax on their first $300,000 of active business income. 7 7 The term "tax expenditure" was coined by United States Assistant Secretary of the Treasury Stanley S. Surrey. Surrey and McDaniel describe tax expenditures as "departures from the normative tax structure [representing] government spending for favoured activities or groups, effected through the tax system rather than through direct grants, loans, or other forms of government assistance." Stanley S. Surrey and Paul R. McDaniel, Tax Expenditures (Cambridge, Mass.: Harvard University Press, 1985). 32 In general, two kinds of tax reduction activities qualify as acceptable tax mitigation. Common to both is that taxpayers obtain a tax reduction in circumstances where the statute did not intend them to be taxed. 7 8 The first type of acceptable tax mitigation involves transactions that result in the economic consequences that Parliament would have expected a taxpayer to have incurred in order to qualify for the favourable tax treatment. Recognizing such transactions as legitimate tax mitigation rests on a view of tax law as a tool of economic policy, and recognizes that normative ideal tax results are intended to be based on the economic outcomes of transactions. Therefore, taxpayers who engage in transactions that appear to qualify for tax reductions, but who have not actually made the corresponding economic sacrifices, should not, as a matter of tax policy, receive those tax reductions. For instance, taxpayers who choose leisure over work in order to avoid higher taxes should not be denied the resulting tax reduction because they have incurred economic sacrifices by foregoing income. Similarly, taxpayers who incur costs to acquire capital assets are permitted to deduct an allowance for the capital cost of that asset from income. 7 9 However, taxpayers who arrange for someone else (such as a spouse) to receive their income but still benefit from it, or taxpayers who purchase capital assets with borrowed money that is immediately repaid so that they assume no economic risk of loss, have not made the 80 economic sacrifices expected to be incurred in order to escape taxation. A s compared to Uckmar, supra note 52 at 23. Uckmar refers to tax mitigation as "tax saving." Also Waincymer, supra note 10 at 251. 7 9 Under paragraph 20(l)(a) of the Act and Regulation 1100 of the Income Tax Regulations, CRC c. 945, as amended [the "Income Tax Regulations"]. 80 Report of the Royal Commission on Taxation, supra note 10 at 542-43. For examples of "circular financing" cases in which borrowed money in respect of which the taxpayer assumed no economic risk of loss was used to acquire capital assets, see Canada Trustco Mortgage Company Ltd. v. The Queen, 2005 D.T.C. 5523 (S.C.C.), affirming 2004 D.T.C. 6119 (F.C.A.), affirming 2003 D.T.C. 587 (T.C.C.) [Canada Trustco]; Barclays Mercantile Business Finance Limited v. Mawson (Inspector of Taxes), [2004] U.K.H.L. 51 (H.L.) [Barclays]; and Peterson v. Commissioners of Inland Revenue, [2005] U.K.P.C. (P.C.) [Peterson]. i 33 genuine tax mitigation, therefore, tax avoidance typically involves only the appearance of an economic sacrifice. As Lord Nolan described it in I.R.C. v. Willoughby: [t]he hallmark of tax avoidance is that the taxpayer reduces his liability to tax without incurring the economic consequences that Parliament intended to be suffered by any taxpayer qualifying for such reduction in his tax liability. The hallmark of tax mitigation, on the other hand, is that the taxpayer takes advantage of a fiscally attractive option afforded to him by the tax legislation, and genuinely suffers the economic consequences that Parliament intended to be suffered by those taking advantage of the option. 8 1 A second type of legitimate tax mitigation occurs when taxpayers engage in tax-reducing transactions in reliance on legislative provisions that specifically authorize, or even encourage, such activities. Permitting taxpayers to take advantage of such provisions recognizes that tax law includes numerous tax expenditures and other incentives enacted specifically to encourage taxpayers to do certain things. In these circumstances, tax minimization is not objectionable because Parliament intends for a tax reduction to occur. This is so even if the activities are highly artificial and do not involve the real economic sacrifice referred to by Lord Nolan in Willoughby above. For instance, taxpayers who make R R S P contributions receive tax deductions for the amounts contributed even though they have not divested themselves of the right to those funds or of the benefit of the return to those savings in the future. 8 2 A tax deduction for R R S P contributions is one of the methods chosen by Parliament to further a policy of encouraging private retirement savings; therefore, making a contribution and claiming the tax deduction should not be considered abusive, even when done solely to reduce tax. 8 3 Similarly, small business owners may choose to 81 I.R.C. v. Willoughby, supra note 3. Also Challenge Corporation Ltd., supra note 3, and Ensign Tankers (Leasing) Ltd. v. Stokes (Inspector of Taxes), [1992] 1 A.C. 655 (H.L.). 8 2 Implemented through subsection 146(5) and paragraph 60(e) of the Act. 8 3 This argument is bolstered by the fact that, even if a taxpayer's contribution was motivated solely by the desire to secure a current tax deduction and the taxpayer did not even contemplate saving for retirement, the 34 incorporate in order to secure the small business deduction, a tax incentive implemented by OA Parliament to encourage the growth of Canadian small businesses. Even the decision to incorporate was not motivated by any of the more common non-tax reasons for incorporating, such as limited liability (i.e., even in a situation where taxpayers change the legal form of their business merely to obtain a tax advantage), there is no sound policy reason to deny the deduction, since Parliament has specifically chosen to reduce the normal rate of taxation in these circumstances in order to achieve other policy goals. As a form of subsidy, these tax expenditure provisions are intended to encourage taxpayers to engage in desirable economic activities. Since tax results should depend primarily on economic circumstances, the reason or purpose for engaging in those activities should not normally disentitle a taxpayer from the benefit of a tax expenditure so long as the taxpayer actually undertakes the 85 relevant activity. Principles of fairness and the rule of law dictate that taxpayers who take advantage of tax incentives offered by Parliament should not be denied the benefit of those incentives, even i f their primary motivation in doing so is to secure the tax deduction. As Lord Millett explained in Peterson: . . . i t was obviously never intended that transactions should be struck down merely because they were influenced by the prospect of obtaining a tax advantage. In many cases, but for the anticipated availability of a tax benefit, the taxpayer would never have entered into the transaction at all. Basic very fact of making such a contribution nonetheless furthers Parliament's policy of encouraging private saving for retirement. Although such an RRSP contribution would be an "avoidance transaction" under the GAAR because it was made solely for tax purposes, it would not be disqualified from a deduction because it would not be considered a misuse or abuse of the legislation. That exception is one way in which Parliament intended to preserve legitimate tax mitigation opportunities like tax incentives when it enacted the GAAR. Supra note 76. 8 5 Normally, the purpose for engaging in a transaction is not considered relevant in determining tax consequences. Refer to the subsequent discussion of purpose under the heading "Common Approaches to Identifying Tax Avoidance - Purpose." 35 features of the tax system...clearly allow for the deliberate pursuit of tax advantage.8 6 Acknowledging that these sorts of transactions should not be viewed as abusive, despite a departure from the normative ideal tax treatment dictated by tax policy, respects the fact that Parliament itself elects to depart from the ideal tax treatment in certain circumstances. D. Common Approaches to Identifying Tax Avoidance Unlike tax mitigation, tax avoidance results in tax reductions that, for policy reasons, should be denied to taxpayers. Interestingly, no consensus exists as to what makes some transactions abusive while others are not. Rather, a number of different formulations or definitions of tax avoidance have been proposed. Cooper identifies three common approaches, which revolve around the notions of form, purpose, and policy. These three approaches are not necessarily mutually exclusive, and they can be used independently or in conjunction with each other. i . Form The form approach focuses on the legal "form" of a transaction. "Form" generally refers to the legal relationships that give legal structure to transactions, through which taxpayers achieve desired economic results (which results are often called the "economic substance" of the transactions). This approach is based on the view that tax law applies tax consequences to transactions by reference to the legal relationships created by taxpayers. For 86 Peterson, supra note 80, para. 36. 8 7 Cooper, supra note 30 at 28-30. 36 instance, the Act imposes different tax consequences on transactions such as sales, transfers, loans, and leases, which are different legal relationships between economic actors (the "forms" of the transactions) who engage in similar underlying economic activities involving the transfer of ownership or possession of property (the "substance" of those transactions). Sometimes it is possible for taxpayers to achieve the same (or a substantially similar) economic result through a different set of legal relationships, thereby altering the tax consequences under the Act. Consequently, taxpayers wi l l often choose to vary the legal form of a transaction in order to qualify for more favourable tax consequences for essentially identical economic results. For instance, a taxpayer intending to purchase a residence might simply obtain a loan from a bank and mortgage the property as security. However, if that taxpayer had funds invested in an income-earning endeavour (such as business equity or investments), he or she could withdraw that equity and use those funds to make the purchase, then borrow money to replace that equity, and mortgage the property as security for that loan. In economic terms, the end result of the two courses of action would be nearly identical: the taxpayer would have a sum of money invested and earning income, the taxpayer would have acquired a property, and the taxpayer would owe a debt to the bank. Under the Act, however, the first course of action would not result in a deduction for interest paid on the loan because the borrowed money was directly used for a personal purpose, while the second course of action would result in such a deduction because the borrowed money was directly used for an no income-earning purpose. Therefore, by structuring essentially the same economic 8 8 Paragraph 20(1 )(c) of the Act provides that a taxpayer may deduct, from income from a business or property, interest paid or payable on "borrowed money used for the purpose of earning income from a business or property." 37 transaction differently, a taxpayer can achieve a different tax result (effectively a tax subsidy 89 for the acquisition of a personal asset). Similarly, a taxpayer winding up a corporation holding retained earnings could draw those profits out of the corporation as dividends; but he or she or could also sell their shares in order to realize the value of the corporation's profits as capital gains, which are taxed at a lower rate than dividends. In economic terms, both courses of action effect a transfer of corporate surplus to the shareholder, but the tax consequences differ due to the different legal relationships involved in the two transaction. A sale of shares gives the taxpayer the best of both worlds - the economic equivalent of receiving dividends (the receipt of corporate surplus), but the tax benefits of selling shares (a lower rate of tax on capital gains than on dividends because only a portion of the gains are included in income). The choice to extract the corporate surplus by selling shares instead of declaring a dividend is a tax-motivated selection of one legal form of transaction over another to achieve essentially the same economic outcome, and is commonly known as "surplus-stripping." 9 0 The form approach to tax avoidance concludes that abusive tax avoidance occurs when taxpayers obtain an economic result through artificial legal forms chosen or structured in order to obtain tax consequences that would not otherwise be enjoyed i f the economic outcome were achieved in a more natural, direct manner. 9 1 Subscribing to this approach, Brooks and Head describe abusive tax avoidance as "contrived and artificial schemes which do not change the substantive character of an activity or transaction but may serve 8 9 The facts of this scenario are based on Singleton v. R., 2001 D.T.C. 5533 (S.C.C.) [Singleton]. 9 0 There have been numerous surplus-stripping cases over the years. One example is McNichol v. The Queen, 97 D.T.C. 111 (T.C.C.) [McNichol], in which the Tax Court applied the GAAR to deny the tax reduction resulting from a sale of shares instead of the receipt of dividends. Hill, supra note 56, is another example of a taxpayer recasting an economic outcome (compound interest) into another legal form (simple interest) in order to secure a tax reduction. 9 1 Vanistendael, supra note 52 at 132. 38 nonetheless to bring the activity within some tax-exempt or more tax-favoured legal 92 category." One strength of the form approach to tax avoidance is that it is consistent with the nature of tax law as a legal instrument intended to apply tax consequences to underlying economic outcomes. The form approach identifies the economic outcome of a transaction and concludes that tax avoidance occurs when the taxpayer secures that outcome (or its economic equivalent) but avoids the normative tax treatment of that economic outcome intended by Parliament. As a result, the form approach to tax avoidance supports the achievement of underlying fiscal and economic policies because it ensures that tax is levied in accordance with those policies, which themselves seek to impose tax consequences based on the economic outcome, not the legal form, of transactions. On the other hand, the form approach has some drawbacks. First, taxpayers faced with choosing between alternate legal forms for their transactions should not be forced to choose the higher-taxed alternative so long as there are valid business reasons for choosing a lower-taxed course of action. In fact, one would not expect economically rational taxpayers making business decisions not to factor tax consequences into their decision-making. 9 4 Taxpayers wi l l naturally choose lower-taxed courses of action that achieve the same economic result as higher-taxed ones. Second, 9 2 Brooks & Head, supra note 22 at 71. Brooks & Head also observe that, through "minor or essentially cosmetic changes," taxpayers may achieve "quite massive" tax savings without incurring significant costs, whether in legal fees or in economic terms. Also, Uckmar, supra note 52 at 25, discusses an "economic theory" according to which when taxpayers, "exploiting the formalism and loopholes of the law, [achieve] an economic result equivalent to that which the legislator intended to tax, the law or the Courts should consider it taxable." 9 3 Ideally, the legal forms through which taxpayers realize economic gains or incur economic sacrifices would not affect the tax treatment of those gains or sacrifices. For instance, in an income tax system using a perfect Haig-Simons comprehensive income tax base, the form in which a taxpayer's economic well-being increases, whether dividends or capital gains, would not affect their liability to tax. 94 See Inland Revenue Commissioners v. Brebner, [1967] 2 A.C. 18(H.L.) [Brebner], per Lord Upjohn, who commented that "no commercial man in his sense is going to carry out a commercial transaction except upon the footing of paying the smallest amount of tax that he can." Brebner has been taken as articulating a "choice doctrine," discussed below under the heading "Common Approaches to Identifying Tax Avoidance - Purpose." In the United States, see Magneson v. Commissioner of Internal Revenue, 753 F.2d 1490 (9th Cir. 1985). 39 the mere fact that taxpayer have chosen lower-taxed legal forms does not automatically mean that tax avoidance has occurred. 9 5 Such a choice could be justified for business reasons, even if that legal form is artificial and contrived. Third, the legislation itself often encourages taxpayers to use artificial and contrived legal forms to achieve particular economic outcomes by suspending the normal tax consequences of those outcomes. 9 6 Taxpayers taking advantage of such incentives should not be viewed as engaging in abusive tax avoidance merely because they have chosen contrived, tax-preferred legal forms for their economic activities. They are merely engaging in permissible tax mitigation. i i . Purpose The second common approach to defining tax avoidance focuses on the taxpayers' purposes in undertaking their transactions. This approach holds that, where taxpayers engage in transactions for the purpose of reducing their tax liability, the transactions are abusive and should not be given effect for tax purposes. The purpose-oriented approach is premised on the view that tax law is intended to apply to real economic activities, not to provide taxpayers with an incentive to engage in artificial transactions without real economic substance merely for tax purposes. For instance, a profitable company might transfer its business to a related company with accrued tax losses, and be appointed agent to run the business on behalf of the loss company. The profit would then legally belong to the loss company and could be offset 9 5 Cooper, supra note 30 at 30. 9 6 For instance, divisive reorganizations under subsection 55(2) of the Act and corporate reorganizations using rollover provisions are among the most artificial and contrived transactions in Canadian tax law, but they are specifically permitted by Parliament in order to enable taxpayers to reorganize their business holdings without incurring immediate tax consequences. See Brian Arnold & James Wilson, "The General Anti-Avoidance Rule - Part II" (1986) 36:5 Can. Tax J. 1123 at 1144 [Arnold & Wilson, "Part II"]. 40 by its tax losses. The purpose-oriented approach would not permit the profitable company to benefit from a transaction that allowed it to operate its business under the umbrella of a related company because the sole purpose of the transfer was to reduce the taxes that the profitable company otherwise would have paid. Given the complexity of taxpayers' commercial affairs, in which business decisions are routinely influenced by tax considerations, this approach naturally raises the question of to what degree tax must be a motivating factor for transactions to qualify as tax avoidance. Must the transaction be solely tax-motivated, or is it sufficient for it to be partly so? If "purpose" is indeed an appropriate criteria for identifying tax avoidance, surely a transaction with no purpose other than to obtain a tax benefit would be abusive. On the other hand, it would not necessarily be abusive for a taxpayer pursuing an ultimately non-tax objective (such as a business, investment or family purpose) to choose a less-taxed means of doing so. Some writers advocate a "choice" doctrine, under which taxpayers are entitled to choose the 98 lowest-taxed options for achieving their ultimate business goals or purposes. The House of Lords articulated this doctrine in Brebner, where Lord Upjohn observed: ...when the question of carrying out a genuine commercial transaction...is reviewed, the fact that there are two ways of carrying it out - one by paying the maximum amount of tax, the other by paying no, or much less, tax - it would be quite wrong, as a necessary consequence, to draw the inference that in adopting the latter course, one of the main objects is. . .avoidance of tax." Such a transaction occurred in Stubart, supra note 31. 9 8 Baker, supra note 68. 99 Brebner, supra note 94. Also I.R.C. v. Wesleyan & Gen. Assur. Soc. (1948), 30 Tax Cas. 11, at 25 (H.L.). This issue permeated two recent Canadian cases - Shell Canada Ltd. v. The Queen, 99 D.T.C. 5669 (S.C.C.) [Shell], and The Queen v. Canadian Pacific Ltd., 2002 D.T.C. 6742 (F.C.A.) [Canadian Pacific]. Both cases involved weak currency borrowing arranged to result in advantageous tax treatment as compared to direct borrowing of US and Canadian dollars, respectively. Canadian Pacific decision considered, in the context of the GAAR, whether the ultimate business purpose of borrowing capital saved the tax-motivated borrowing step in the transaction, and concluded that the choice to borrow capital through an artificial transaction that was more tax-efficient than a direct transaction was not abusive. The mechanics of weak currency borrowing transactions are discussed further in Chapter 4. Also refer to Chapters 5 and 6 for discussion of how the GAAR should apply to those transactions. 41 On the other hand, the mere presence of minimal non-tax considerations should not permit taxpayers to benefit from transactions that are primarily tax-motivated, or which would not have been undertaken but for their tax aspects. To be workable, therefore, a purpose-oriented approach should set the non-tax purpose requirement fairly high, such as by requiring tax to be the predominant or primary purpose, not merely an ancillary purpose, before considering a transaction to be tax avoidance. Such a threshold would recognize that the tax system is not intended to provide reasons for taxpayers to engage in transactions that would not otherwise be undertaken for non-tax reasons, while respecting their right to pursue valid non-tax objectives in tax-efficient ways. A distinct advantage of a purpose-oriented approach is that, in the tax avoidance context, the purposes of taxpayers' transactions are generally much easier to identify, based on objective evidence, than the underlying policy or "object and spirit" of the relevant legislative provisions. 1 0 0 In addition, the business purpose test, a form of purpose-oriented approach adopted by the United States Supreme Court in 1935, 1 0 1 has been the primary anti-avoidance rule employed in the United States and appears to have had some effect in controlling tax avoidance. 1 0 2 From a tax policy perspective, the main advantage of the purpose-oriented approach is that it acknowledges that tax law is intended to be as Brian Arnold, "In Praise of the Business Purpose Test," in Report of the Proceedings of the Thirty-Ninth Tax Conference, 1987 Conference Report (Toronto: Canadian Tax Foundation, 1988) 10:1-34, at 10:10 [Arnold, "Business Purpose Test"]. Also John Tiley, "Judicial Anti-Avoidance Doctrines: The US Alternatives - Part II" [1987], no. 6 British Tax Review 220 at 220 [Tiley, "Part II"]. 101 Gregory v. Helvering, 293 U.S. 465 (1935), affirming Helvering v. Gregory, 69 F.2d 809 (2d Cir. 1934) [Gregory v. Helvering]. 1 0 2 It is impossible to empirically track the success or failure of an anti-avoidance rule because it is impossible to know, other than anecdotally, whether taxpayers' subjective decision-making is influenced by the potential application of anti-avoidance rules. Similarly, it is impossible to determine the degree to which tax avoidance is not deterred by an anti-avoidance rule if it goes undetected by the tax authorities. However, the fact that the business purpose test has been regularly and consistently applied in reported U.S. cases suggests that it has in fact had some success as an anti-avoidance rule. 42 economically neutral as possible. Although tax considerations wi l l inevitably influence economic and commercial decision-making, tax alone should not be the reason for taxpayers' undertaking transactions. When taxpayers engage in transactions that they would not otherwise consider, merely to obtain tax benefits, therefore, the spectre of abuse arises. However, the purpose-oriented approach also has some weaknesses. One is that, from a tax policy perspective, taxpayers' motives and purposes are generally irrelevant to the taxation of their transactions because tax consequences are typically intended to result more from objective economic circumstances (such as whether property was sold or an expense was incurred) than from taxpayers' purposes for their transactions - except for in specific situations where legislation may make purposes relevant, such as when distinguishing 103 between business and personal expenses. A more significant problem with the purpose-oriented approach relates to the numerous tax expenditure provisions that encourage certain activities by granting tax reductions at the expense of tax law's primary purpose of raising revenue. Some taxpayers w i l l likely engage in certain transactions in order to obtain those tax incentives even without an underlying non-tax purpose. 1 0 4 A s explained above, a taxpayer may choose to contribute to an R R S P solely (or even primarily) to obtain a current tax deduction, not because their purpose is to save for retirement. If "purpose" distinguishes between tax avoidance and tax mitigation, then two taxpayers making identical R R S P Paragraph 18(l)(a) of the Act permits deductions from business or property income only when expenses are incurred "for the purpose" of producing income from a business or property. However, from a purely tax policy perspective, it might be possible to distinguish business from personal expenses simply by considering the economic effect of the expenditure, and whether it relieved the taxpayer of a personal expenditure that would otherwise have been incurred. 1 0 4 This problem was identified by the Supreme Court of Canada in Stubart, supra note 31 at 6322, as a reason not to adopt a general business purpose test in Canadian tax law. The Court went so far as to speculate that, where Parliament has offered a tax incentive, taxpayers might choose to undertake the activity that qualifies for the tax incentive without any business purpose, and that requiring a business purpose in order to qualify for the tax reduction might actually discourage or even bar taxpayers from undertaking the very activity that Parliament sought to encourage. Thus, the Court suggested, the decision to offer a tax incentive could perversely have a deterrent effect on taxpayers' behaviour. 43 contributions, where one genuinely intended to save for retirement and the other sought only to obtain a tax deduction, would be treated unequally because of their different purposes. Such a situation would be a failure of horizontal equity, which dictates that two taxpayers making identical R R S P contributions should be taxed identically. Furthermore, it is not logically possible to treat only some of the various tax incentive provisions in this fashion. If we accept the equity argument that taxpayers making R R S P contributions solely for tax reasons should receive that deduction, then we cannot logically justify denying any taxpayer who complies with the requirements for any other tax incentive the benefit of that incentive, regardless of their purpose in doing so, since the entitlement to incentives normally depends on the objective economic circumstances (such as an R R S P contribution having been made) more than on the taxpayer's purpose. Thus, the purpose-oriented approach alone is not satisfactory, either. i i i . Policy The third main approach to identifying tax avoidance focuses on the policies underlying the legislation. Under this approach, tax avoidance occurs when taxpayers obtain tax consequences not intended by Parliament or by the scheme of the legislation, or when tax the results of their activities defeat the policy underlying the legislation. This approach has been articulated in a number of different ways. For instance, the U K Tax Law Review Committee described tax avoidance as any "action taken to reduce or defer tax liabilities in a way that Parliament plainly did not intend or could not possibly have intended had the matter 44 been put to it. Parsons, an Australian author, describes tax avoidance as an "incongruence" between the tax outcome and the government's purpose that results when the legislation is mechanically applied without attempting to give effect to its underlying purpose or po l i cy . 1 0 6 Hogg, a Canadian author, suggests that tax avoidance becomes abusive when taxpayers are " in conflict with the general policy of the statute, and [are] taking advantage of 107 faulty or incomplete legislative language which has been used to give effect to the policy." The policy-based approach is premised on the idea that tax law does have some underlying policy aims, and that taxpayers' transactions can sometimes lead to results that run counter to that policy, particularly where the legislation as it is drafted does not anticipate the manner in which the taxpayer uses the provisions. Imperial Oil v. The Queenm is an example of a transaction creating tax results running counter to the policy of the applicable legislation. Imperial Oil concerned Part 1.3 of the Act, which applies a flat rate tax to the capital of a large corporation 1 0 9 at the corporation's year-end (a form of corporate capital tax). Imperial O i l made short-term loans totalling $500 mil l ion over its year-end to other corporations in order to be able to include the principal of those loans in its "investment allowance," an amount deducted from its taxable capital in computing Part 1.3 tax. 1 1 0 The investment allowance deduction is based on the assumption that the borrower would include the principal of the loans in its own capital for Tax Law Review Committee, supra note 14 at 3. 1 0 6 R.W. Parsons, Income Taxation in Australia (Sydney: Law Book Co., 1985), at paras. 16.55-16.58, cited in Cooper, supra note 30 at 30. 1 0 7 Peter Hogg, Joanne Magee & Jinyan Li, Principles of Canadian Tax Law, 4th ed. (Toronto: Carswell, 2002), at 554-55. 108 Imperial Oil Ltd. v. The Queen, 2004 D.T.C. 6044 (F.C.A.), affirming 2002 D.T.C. 1954 (T.C.C.) [Imperial Oil]. Both the Tax Court and the Federal Court of Appeal held that the GAAR did not apply to these transactions. 1 0 9 Under subsection 181.1(1) of the Act, Part 1.3 tax is payable only by corporations with capital of greater than $10 million. ' 1 0 The investment allowance is computed under subsection 181.2(4) of the Act and subtracted from the corporation's taxable capital under subsection 181.2(2). 45 purposes of Part 1.3 tax. The investment allowance is therefore intended to prevent double taxation of corporate capital. However, one of Imperial Oi l ' s borrowers had a different year-end date, so that when it calculated its capital for Part 1.3 purposes, the funds had already been returned to Imperial O i l . The result was that neither corporation paid Part 1.3 tax on the capital, and the capital escaped Part 1.3 tax completely. 1" Imperial Oi l ' s transaction clearly exploited a loophole in the legislation, which did not anticipate loans between corporations with different year-ends. The underlying policy of Part 1.3, which is intended to generally tax the capital of large corporations carrying on business in Canada, was thereby defeated. Under a policy-oriented approach, therefore, this transaction would be tax avoidance. From a tax policy perspective, the advantage of this approach is apparent. It focuses on the fiscal and economic policies underlying the legislation, and measures transactions against that policy. This approach recognizes that, with respect to a given transaction, it should be possible to identify a normative ideal tax treatment intended by Parliament, or one which would have been intended by Parliament, even i f the legislation does not specifically contemplate the given transaction. In addition, this approach allows for both types of tax mitigation because it acknowledges that some tax-reduction transactions actually accord with the purpose of the legislation. 1 1 2 On the other hand, the disadvantage of this approach is that it can sometimes be extremely difficult to clearly identify a policy underlying tax legislation. It is arguable that "Parliamentary intent" is merely a fiction in the tax context, since it is highly unlikely that members of Parliament clearly understand the detailed legislative '"in addition, another of Imperial Oil's borrowers was not a Canadian company. To the extent that that borrower did not carry on business in Canada, the borrowed funds would also have escaped tax under Part 1.3 because tax applies only to "taxable capital employed in Canada," as defined in subsections 181.1(1) and 181.2(1) of the Act and Regulation 8601 of the Income Tax Regulations. It is arguable, however, that loans to non-residents not employing the capital in Canada would not be considered tax avoidance under a policy-oriented approach, since the legislation clearly intends to tax only capital employed in Canada. 1 1 2 Cooper, supra note 30 at 31. 46 provisions that they enact, or why they do so. Only personnel from the Department of Finance, who draft Canadian tax legislation, are likely to have a comprehensive understanding of all of the different policies expressed in the legislation. Moreover, complex cases could involve competing policy objectives, such as raising revenue on one hand and encouraging the accumulation of capital through capital cost allowance or interest deductions on the other. How is a court to identify and balance competing policies, and to determine 113 which policies should prevail in a particular case? This question is particularly problematic because courts are used to applying tax law based on its words, not to considering underlying tax policy more generally. Although the policy approach to tax avoidance is good tax policy, it could prove impractical by requiring the Courts to "go behind" the legislation to deal with questions of underlying policies that Courts are not well suited to answering. 1 1 4 E. A Tax Policy-Based Theory of Tax Avoidance Thus, each of the three main approaches to identifying tax avoidance has strengths and weaknesses. None is adequate on its own. I suggest that a coherent, tax policy-oriented approach to identifying tax avoidance should draw on elements of all three approaches, while 1 1 3 David Ward reviews some examples of the difficulty of identifying underlying policy considerations, which he describes as the "object and spirit" of particular provisions of the Income Tax Act, in "Tax Avoidance: Judicial and Legislative Approaches in Other Jurisdictions," in Report of the Proceedings of the Thirty-Ninth Tax Conference, 1987 Conference Report (Toronto: Canadian Tax Foundation, 1988), 8:1 at 8:16-17 [Ward, "Judicial and Legislative Approaches"]. Arnold, "Business Purpose Test", supra note 100 at 10:7, argues that the concept of the "object and spirit" of the Income Tax Act is "so vague and ambiguous as to be virtually meaningless." 1 1 4 David A. Ward et al. sound a cautionary note about the appropriateness of courts engaging in this exercise, suggesting that a court merely "[applies] its intuition to frustrate what it perceives to be a tax avoidance transaction" and may not be able to correctly identify the intention of the legislature, particularly given the complex web of economic incentives, tax benefits, and tax expenditures. See Ward et al., "The Business Purpose Test and Abuse of Rights" [1985], no. 1 British Tax Review 68 at 121 [Ward, "Business Purpose Test"]. 47 retaining its focus on underlying issues of tax policy. Brooks offers one such integrated theory, which includes the following three elements: (a) tax avoidance transactions "[result] in a mismeasurement of taxpayers' economic income so that they pay less tax than they would have paid if they were taxed on their economic income;" (b) tax avoidance transactions are "engaged in by taxpayers for the sole or primary purpose of obtaining such a tax benefit;" and (c) the mismeasurement of economic income created by tax avoidance transactions is "not a mismeasurement...that was contemplated for administrative or policy reasons by the structure of the tax legislation."" 5 From a tax policy perspective, Brooks' formulation is generally sound. B y recognizing that income (the tax base) is an economic concept and that tax law is no more than a series of rules for measuring that economic concept in legal terms, he recognizes that tax law is a policy instrument intended to implement underlying economic and fiscal policies. He also recognizes that, because tax law attempts to measure an economic concept using legal concepts, that measurement wi l l not always be correct. This point is closely related to the form approach, which suggests that tax avoidance results when taxpayers vary the legal form of their transactions in order to qualify for different tax treatment for essentially similar economic outcomes. I believe that this theory would be improved, however, by amending paragraph (a) above to provide that a tax avoidance transaction is one where, because of its application to the legal form of the taxpayer's transaction, the legislation fails to impose the tax consequences that, from a tax policy perspective, would have been intended to apply to 1 1 5 Neil Brooks, "The Responsibility of Judges in Interpreting Tax Legislation" in Cooper, Tax Avoidance and the Rule of Law, supra note 10, 93 at 96. Joshua D. Rosenberg, "Tax Avoidance and Income Measurement" (1988-1989) 87 Mich. L. Rev. 365, also discusses the concept of an "undermeasurement" of economic income, and writes at 447 that "tax avoidance exists only when there is a convergence of undermeasurement of economic income with a failure to achieve the specific goals underlying the provision that allows for that undermeasurement." 48 the economic outcome of the transaction. Such a formulation more clearly acknowledges that tax avoidance lies in the failure of the legislation, applying as it does to the legal relationships and forms of a taxpayer's transaction, to impose the consequences that tax policy dictates from an economic perspective. Thus modified, this theory is a sound, policy-based approach to defining tax avoidance. It acknowledges that tax avoidance lies in the failure of the legislation to implement the normative ideal tax results dictated by tax policy. It also acknowledges that tax avoidance becomes abusive when transactions are undertaken solely or primarily for the purpose of obtaining a tax reduction, not for business or other valid non-tax purposes. This element is consistent with the general presumptions that favour the rights of taxpayers to arrange their affairs and to anticipate the tax results of transactions in advance. It is also consistent with the important objective of not unduly interfering with commercial decision-making. So long as taxpayers are pursuing valid non-tax objectives, decisions to do so in tax-efficient ways wi l l not automatically be viewed as abusive. Tax avoidance wi l l occur only when taxpayers structure their transactions in order to minimize the tax consequences of pursuing a non-tax objective in a manner that frustrates or defeats the policy aims underlying the legislation. Finally, this theory acknowledges that there are instances in which tax policy requires a departure from the normative ideal tax results, such as in the case of incentive provisions in which Parliament has consciously sacrificed tax revenue in pursuit of other policy goals. This theory focuses the analysis of tax avoidance where it should be - squarely on policy. In the final Chapter, after a comparative review of various approaches to controlling tax avoidance and an analysis of the current G A A R , this thesis w i l l test whether the theory developed in this Chapter can be applied two real tax avoidance scenarios - a 49 circular financing sale-leaseback and surplus stripping. That Chapter w i l l explore how the current G A A R can be applied in accordance with this theory and in a manner that preserves normative ideal tax outcomes in tax avoidance situations. 50 IV. COMPARATIVE APPROACHES TO CONTROLLING TAX AVOIDANCE Because tax avoidance creates serious problems for the tax system, as outlined in Chapter 2, tax lawyers and academics studying tax avoidance have struggled mightily to determine how the legal system should control it. Because tax law, even though fundamentally a policy instrument, is nonetheless law, any rules for identifying and preventing tax avoidance should be based on rational, systematic criteria. In light of the theoretical formulation of tax avoidance developed in the previous Chapter, the challenge is to preserve normative ideal tax outcomes when taxpayers undertake transactions that create legal relations giving rise to tax results that, for policy reasons, are not the results that Parliament would have intended to apply to the underlying economic outcome of those transactions. 1 1 6 This Chapter reviews a number of legal doctrines that are applied by tax authorities and courts in various jurisdictions to rationalize departing from the tax results that would otherwise apply to transactions based on their form and legal relationships, in order to impose tax consequences based on overarching considerations of tax policy. A s John Tiley observes, the problems encountered in developing rational guidelines for determining when to do so are "timeless and insoluble." 1 1 7 Most Western jurisdictions, regardless of the nature of their legal system, face a similar tax policy problem, and a great deal of intellectual energy 1 1 6 As explained in Chapter 3, normative ideal tax outcomes are the tax consequences that Parliament intended, or would have intended, to apply to the outcome of a transaction. Unless, as in the case of tax incentive provisions, Parliament clearly intends otherwise, normative ideal tax consequences are determined primarily by a taxpayer's economic circumstances. For examples of how normative tax consequences can be identified and realized in tax avoidance cases, refer to Chapter 6. 1 1 7 John Tiley, "Judicial Anti-Avoidance Doctrines: The US Alternatives", [1987] no. 5 British Tax Review 180 at 185 [Tiley, "The US Alternatives"]. 51 has been spent considering, reviewing, and arguing for and against various approaches to controlling tax avoidance. 1 1 8 For the most part, Western jurisdictions have relied on a combination of judicial doctrines and statutory provisions to control tax avoidance. Statutory provisions can be either detailed rules aimed at preventing specific tax avoidance transactions; 1 1 9 or wide-ranging rules of general application that override the literal effect of the legislation to redetermine tax results in keeping with the scheme of the legislation or other policy considerations. 1 2 0 B y contrast, anti-avoidance doctrines are not typically found within the four corners of a statute, but are applied as interpretive aids in the context of deciding tax disputes. Although the problems associated with properly identifying tax avoidance are common, each jurisdiction has adopted its own particular mix of solutions in the unique context of its domestic legal system. In addition to specific, targeted anti-avoidance provisions - which necessarily are only of limited value because they react to, rather than anticipate, tax avoidance activities - Canada has elected to rely primarily on a statutory general anti-avoidance rule. Parliament elected to enact such a statutory rule primarily due to concerns that Canadian courts were unwilling to vigorously apply anti-avoidance doctrines and that specific anti-avoidance provisions alone were inadequate to properly deal with tax avoidance. For a useful review of the various approaches taken in Canada, the United States, the United Kingdom, Australia, and other Western European countries, refer to Arnold & Wilson, "Part I", supra note 74, and Vanistendael, supra note 52. 1 1 9 For examples of some Canadian targeted anti-avoidance rules, refer to section 20.3 of the Act, which enacts rules aimed at weak currency loans, and to section 120.4, which enacts "kiddie tax" rules aimed at preventing income-splitting by diverting income to a minor child. 1 2 0 The GAAR is one such provision. Many other jurisdictions, such as Australia, New Zealand, and France have also adopted such general statutory provisions. 52 A. The Importance of Statutory Interpretation in Combating Tax Avoidance Generally, successful tax avoidance relies on using the literal meaning of legislative provisions in ways that the legislature did not anticipate or intend. Therefore, in order to strike down a tax avoidance transaction, the literal effect of the legislation must be disregarded and some other interpretation must be given effect. In such a situation, the courts' approach to statutory interpretation is of primary importance because the interpretive exercise is concerned with how to apply a legislative provision to a set of facts, taking into account considerations such as the meaning of the provision, its context in the overall scheme of the legislation, and the purpose of the provision. 1 2 1 A court deciding a tax avoidance case must choose between competing interpretations of the relevant provisions when would result in quite different tax consequences. Therefore, different approaches to statutory interpretation affect how restrictive or permissive the legal climate is with regard to tax avoidance. A n approach that gives more weight to the underlying policy goals of Parliament, as it is expressed in the overall structure and scheme of the legislation, w i l l be less likely to permit taxpayers to benefit from tax avoidance because tax avoidance typically requires a literal application of the legislation without regard to overarching tax policy considerations. On the other hand, an approach that focuses on a literal interpretation of the legislation, but 1 2 1 The definitive statement of the "modern rule" of statutory interpretation is formulated by Driedger: ".. .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." E.A. Driedger, Construction of Statutes, 2d ed. (Toronto: Butterworths, 1983) at 87. The Supreme Court of Canada endorsed this rule of interpretation in tax law in Stubart, supra note 31. In addition to the factors listed by Driedger, Duff suggests that the consequences of alternative proposed interpretations is also a relevant factor in the interpretive exercise. David Duff, Canadian Income Tax Law (Toronto: Edmond Montgomery Publications Limited, 2003) at 73-74. Taylor argues that Canadian courts, following the lead of the Supreme Court of Canada, have resisted seriously considering the tax policy consequences of alternative interpretations in tax avoidance cases. Roger Taylor, "The Supreme Court of Canada: Principles of Adjudication of Tax-Avoidance Appeals from Stubart to Shell Canada," in Report of the Proceedings of the Fifty-First Tax Conference, 1999 Conference Report (Toronto: Canadian Tax Foundation, 2000), 17:1-53 at 17:7. 53 which does not consider the consequences of a literal interpretation in terms of normative policy considerations, w i l l be more likely to permit tax avoidance. In general, if courts deciding tax disputes focus simply on determining the meaning of the relevant legislative provisions as opposed to evaluating whether competing interpretations either achieve or frustrate the intent of Parliament, their interpretation wi l l be more permissive toward tax avoidance. In addition, how courts perceive their role in the adjudication of tax disputes, and in particular whether courts acknowledge any "supervisory responsibility" with respect to the integrity of the tax system as a whole, is another important influence on their response to tax avoidance. 1 2 2 In most Western jurisdictions, such as the United States, the U K , and the major c ivi l law countries of Western Europe, courts have developed a variety of judicial anti-avoidance doctrines to stop tax avoidance. Generally, those anti-avoidance doctrines provide a framework for rationalizing when to deny taxpayers the beneficial consequences resulting from a literal reading of the legislation in circumstances where no statutory provisions mandate the court to do so. Where courts have adopted a more policy-oriented approach to statutory interpretation, they have developed more vigorous anti-avoidance doctrines. Where, as in Canada, courts have taken a more restrained approach to statutory interpretation and have not embraced inquiries into broad tax policy considerations, they have developed only restricted anti-avoidance doctrines. The most common anti-avoidance doctrines developed in Western jurisdictions are the sham transaction doctrine, the ineffective transaction doctrine, the business purpose test, 1 2 2 The importance of the court's view of its role is discussed both by Taylor, ibid., and by Arnold, "Reflections", supra note 10. Refer to Chapter 5 for discussion of how the Canadian courts perceive their role in deciding tax cases. 1 2 3 This point is advocated by Taylor, supra note 121 at 17:7, and by Arnold, ibid. the step transaction doctrine, the substance over form doctrine, and the abuse of rights doctrine. 1 2 4 Each of these doctrines attempts to provide a rational justification for ignoring the literal application of the legislation to a tax avoidance transaction. Rather, they attempt to interpret the legislation in a manner more consonant with the legislative intent underlying the relevant statutory provisions. 1 2 5 However, no jurisdiction employs all of these doctrines, nor are the doctrines themselves necessarily consistent from one jurisdiction to another. For instance, the sham transaction doctrine has historically been formulated differently in the United States as compared to in Canada and the U K . Similarly, the abuse of rights doctrine is known only in c iv i l law jurisdictions (although the G A A R appears to have been an attempt to import the concept of "abuse" into Canadian law). In this Chapter, I argue that most of the various anti-avoidance doctrines employed in other jurisdictions have been rejected by Canadian courts. Only the sham transaction and ineffective transaction doctrines survive as valid doctrines, but, in my view, neither may properly be considered an anti-avoidance doctrine. Canadian courts have been wedded to a traditional, literal approach to statutory interpretation, focusing on the straightforward application of legislative provisions to the legal form of taxpayers' transactions without regard for overarching policy considerations. After reviewing the various extra-statutory anti-avoidance doctrines identified above, I conclude that the failure of Canadian law to develop vigorous anti-avoidance doctrines may be traced directly to an unsatisfactory approach to the interpretation of tax statutes taken by the Supreme Court of Canada since Stubart. That approach has been characterized by renewed adherence to literal interpretation 1 2 4 In this Chapter, I argue that the sham transaction and ineffective transaction doctrines are not true anti-avoidance doctrines because they would not apply to true tax avoidance transactions. 1 2 5 Except for the sham transaction and ineffective transaction doctrines. Those doctrines do not actually ignore the literal application of legislative provisions, but rather outright deny that a taxpayer is entitled to rely on the legislative provisions in issue. 55 and a reluctance to accept that courts have a necessary role to play in controlling tax avoidance. Rather, that approach consistently characterizes tax avoidance as a problem for Parliament alone to remedy. B. The Sham Transaction Doctrine In Canada, the U K , and the United States, a transaction is ignored for tax purposes if it is a sham. However, the meaning of "sham" is not the same in the three jurisdictions. In the U K , the concept of sham is confined to transactions in which "the label given by the parties is not conclusive in determining the legal rights created by the parties." 1 2 6 A transaction is a sham where the parties claim to have created legal rights that, in law, they have not actually created. The classic U K description of a sham is found in the judgment of Diplock J. in Snook v. London & West Riding Investments Ltd.: . . . [a sham] means acts done or documents executed by the parties to the "sham" which are intended by them to give to third parties or to the court the appearance of creating between the parties legal rights or obligations different from the actual legal rights and obligations (if any) which the parties intended 127 to create. In English tax law, the central element of a sham is deceit: a transaction is a sham only when the parties, in their documents, have merely created the appearance of rights that they did not actually intend to create. 1 2 8 The purpose of a sham transaction is to induce third parties, such as the tax authorities, to treat the transaction as i f certain legal relationships existed when in reality they do not. In the tax avoidance context, a transaction is a sham when the parties purport to create particular legal relationships for the purpose of qualifying for beneficial tax 1 2 6 Tiley, "The US Alternatives", supra note 117 at 196. 1 2 7 [1967] 1 All E.R. 518 at 528 (C.A.) [Snook]. 128 Stubart, supra note 31 at 6021. Also Arnold & Wilson, "Part I", supra note 74 at 855. 56 results but have not, in fact, done so. Because sham transactions do not actually create the legal relations that they purport to create, shams are ignored for tax purposes. Instead, tax consequences are imposed on the basis of the actual legal relations created by the parties, despite the apparent creation of other legal rights. Furthermore, for English tax law purposes, whether the parties to a transaction have incurred any economic price (in other words, whether there is a change in the parties' underlying economic circumstances) to create the apparent legal relationships is not relevant. The sham doctrine deals with transactions only in terms of their legal form, not their economic substance. For that reason, the sham transaction doctrine is not concerned with ensuring the normative tax outcomes that Parliament would have intended to apply to a transaction. B y contrast, in the United States the sham doctrine is applied more widely than in the U K . In American tax law, the sham doctrine is neither confined to situations of deceit (a limitation which, as Tiley observes, makes a sham tantamount to an attempt at fraud ), nor is it concerned solely with the legal form of transactions as opposed to their underlying economic substance. Rather, the American concept of sham refers generally to transactions in which a legally-effective transaction does not have any underlying economic substance apart from the creation of tax losses. Accordingly, in the United States, the legal form of a sham transaction is disregarded in taxing the transaction. Therefore, the American sham doctrine is a form of legal shorthand evidencing a general preference for taxing in accordance 1 2 9 Arnold & Wilson, "Part I", ibid, at 854. 1 3 0 Tiley, "The US Alternatives", supra note 117 at 196. 131 Ibid, at 196. Also Jacobson v. Commissioner, 915 F.2d 823, 837 (2d Cir. 1990) and DeMartino v. Commissioner, 862 F.2d 400, 406 (2d Cir. 1988). For a competing view that the sham doctrine ought to be limited to instances where taxpayers pass off transactions as something other than they actually are and that any further analysis of a transaction's underlying substance improperly conflates supposedly distinct sham and economic substance doctrines, refer to Stewart Patton, "Treasury Regulation § 301.6111-2T and the Economic Substance Doctrine: A Plea for Certainty in the Tax Law" (2002-2003) 39 Hous. L. Rev. 499 at 506-07. 57 with the economic substance of transactions as opposed to their legal form, a preference which is not found in U K or, for that matter, Canadian, tax law. In Canada, the sham doctrine has traditionally closely paralleled the narrow U K formulation that limits a sham to virtually fraudulent situations. In the 1970s, however, the courts briefly expanded the sham doctrine in M.N.R. v. Leon. In Leon, the court characterized a transaction as a sham merely because it lacked a business purpose. 1 3 3 In doing so, the Leon decision expanded the sham doctrine beyond the confines of the U K approach and brought the Canadian approach closer to the US approach, although it did not explicitly incorporate considerations of economic substance. However, Canadian courts soon retreated from the Leon formulation in Massey-Ferguson Ltd. v. The Queen,134 and, ultimately, in Stubart the Supreme Court of Canada reaffirmed the U K approach to sham, distinguishing the sham doctrine from a business purpose test and limiting shams to cases of deceit. In Stubart, Estey J. confirmed that, in Canadian law, "deceit.. .is the heart and core of a sham." 1 3 5 Post-Stubart, therefore, it is clear that for Canadian tax purposes a sham exists only when the conditions outlined in Snook are met. Specifically, a sham means a transaction that attempts to deceive a third party by creating the appearance of legal relations that have not actually been created. Unlike in the United States, the sham doctrine does not apply to situations in which the legal form of a transaction does not accurately reflect its economic substance. Similarly, the purpose behind a tax avoidance transaction is not relevant to 132 Minister of National Revenue v. Leon et al, 76 D.T.C. 6299 (F.C.A.) [Leon]. 133 Ibid, at 6302. 1 3 4 77 D.T.C. 5013 at 5020 (F.C.A.) [Massey-Ferguson]. 135 Stubart, supra note 31 at 6321. For a brief period after The Queen v. Bronfman Trust, 87 D.T.C. 5059 (S.C.C.) [Bronfman Trust], it appeared that the concept of sham might once again expand to include artificial transactions whose legal form does not reflect their underlying economic outcomes. This suggestion was closely tied to the "substance over form" doctrine, discussed later in this Chapter. However, the Supreme Court of Canada put a definitive end to this possibility in Shell, supra note 99. 58 whether it is a sham. A l l that matters is the legal reality of the transaction. In this way, the Canadian sham doctrine is perhaps the truest manifestation of the principles that the substance of a transaction is to be found in the real legal relationships created by the transaction, not in its underlying economic result, and that a tax-reduction purpose does not render a transaction ineffective for tax purposes. Only a transaction that merely creates the appearance of legal relations that are not actually created wi l l be a sham. In my view, the sham doctrine is a strikingly unsatisfactory anti-avoidance doctrine for two reasons. First, the Canadian sham doctrine is so narrowly defined that very few transactions ever come within the definition of a sham. B y confining shams to situations in which the parties to a transaction intend to deceive a third party like the tax authorities, Canadian courts have limited the sham concept to only the most extraordinary cases, such as Massey-Ferguson.137 Notably, since Stubart there have been no high-profile Canadian tax avoidance cases in which a court has struck down a tax avoidance transaction as a sham. The reason for the scarcity of sham cases is the second, and most important, failing of the sham doctrine as an anti-avoidance doctrine. Most tax avoidance transactions involve the deliberate creation of actual legal relationships, not an attempt to deceive the tax authorities as to the existence of those relationships. Tax avoidance relies on the creation of valid legal relationships that, on a literal application of the legislation, entitle a taxpayer to beneficial tax consequences. Generally, tax avoidance transactions are not shams because the parties to those transactions actually create the legal relationships that they purport to Those are two of the key principles enunciated by the House of Lords in Duke of Westminster, supra note 53. Arnold discusses the central importance of three main Duke of Westminster principles in the history of Canadian tax law, in "Reflections", supra note 10. Those principles are reviewed in more detail in Chapter 5 under the heading "Statutory Interpretation and Tax Avoidance in Canada." 137 Supra note 134. 1 3 8 Arnold & Wilson, "Part I", supra note 74 at 856. 59 create. Furthermore, most tax avoidance transactions are carried out with professional advice from lawyers and accounting advisors who are likely to ensure that those transactions are legally effective. Only the most amateur attempts at tax avoidance are likely to attempt to deceive third parties with regard to their actual legal effect. Most tax avoidance transactions do not need to deceive to succeed, and therefore there have been very few, if any, cases of true shams since Stubart. C. The Ineffective Transaction Doctrine A second anti-avoidance doctrine is known as the "ineffective transaction" 139 doctrine. Tax authorities and courts invoke the ineffective transaction doctrine to strike down tax avoidance transactions on the basis that they are legally incomplete and have not successfully created the legal relations that give rise to the desired tax results. A transaction is legally ineffective where the parties have failed to observe or complete all of the technical requirements of the transaction. The ineffective transaction doctrine differs from the sham doctrine in that there is no requirement of deceit. Rather, it applies when a taxpayer has simply failed to carry out all of the legally necessary steps, whether through inadvertence or otherwise. 1 4 0 The rationale for this doctrine is that a taxpayer should not be permitted to achieve a reduction of tax without having validly established the requisite legal relations to which the legislative provisions apply. A classic Canadian statement of this doctrine is found 139 Ibid, at 852. Also Krishna, supra note 2 at 995 and 1001-02; and Hogg et al., supra note 107 at 566. 1 4 0 In a number of recent cases, such as Juliar, infra note 153, and Snow White, infra note 154, taxpayers have asked courts to rectify transactions that are incomplete or which mistakenly do not reflect the taxpayers' original intentions. Under the heading "Rectification," I suggest that the advent of rectification could mean the end of the ineffective transaction doctrine. 60 in Atinco Paper Products Ltd. v. The Queen: .. .it is the duty of the Court to carefully scrutinize everything that a taxpayer has done to ensure that everything which appears to have been done, in fact, has been done in accordance with applicable law. It is not sufficient to employ devices to achieve a desired result without ensuring that those devices are not simply cosmetically correct, that is, correct in form, but, in fact, are in all respects legally correct, real transactions. If this Court, or any other court, were to fail to carry out its elementary duty to examine with care all aspects of the transactions in issue, it would not only be derelict in carrying out its judicial duties, but in its duty to the public at large.. .The only course for the Court to take is to apply the law as the Court sees it to the facts as found in the particular transaction. If the transaction can withstand that scrutiny, then it w i l l , of course, be supported. If it cannot, it w i l l f a i l . 1 4 1 If a court concludes that transactions are legally ineffective, taxpayers' tax liability wi l l be based on their true legal position. In Stubart, the Supreme Court of Canada recognized the ineffective transaction doctrine as a valid doctrine, although the Court declined to apply it in that case. The ineffective transaction doctrine has not been the subject of much jurisprudence in Canada, partly because it is not a particularly vigorous anti-avoidance doctrine, and partly because it is not much of a departure from the approach taken in regular tax cases. A court properly carrying out its judicial responsibility should always review the legal effectiveness of transactions before affirming taxpayers' entitlement to tax consequences, even i f the cases do not involve tax avoidance. The fact that tax avoidance transactions are tax-motivated may make the tax authorities more likely to closely verify their legal effectiveness, but the tax motivation would not normally be relevant to that inquiry. Therefore, it should not be considered unusual if a court denied beneficial tax results to a legally ineffective tax 1 4 1 78 D.T.C. 6387 at 6395 (F.C.A.). The Court refers to a "duty" to scrutinize what taxpayers have done. It is not clear whether the Court meant that courts have a heightened responsibility to examine the legal mechanics of transactions in the tax avoidance context, or merely that a court always has a duty to ensure that taxpayers' transactions are legally effective before endorsing tax consequences. In light of more recent case law, discussed in Chapter 5 under the heading "Statutory Interpretation and Tax Avoidance," it appears that Canadian courts, if they ever did believe there to be a responsibility to scrutinize tax avoidance transactions more closely than normal transactions, no longer believe this to be the case unless the Crown invokes the GAAR. 61 avoidance transaction, and no special anti-avoidance doctrine should be needed to justify doing so. More importantly, the ineffective transaction doctrine has not played a prominent role in Canadian tax law because tax avoidance typically involves properly executed transactions that otherwise qualify for beneficial tax treatment. 1 4 2 There are very few tax avoidance cases in which taxpayers have lost because they did not complete their documentation correctly. Tax avoidance is not a vexing problem because the law has difficulty determining whether or not transactions are legally effective; rather, tax avoidance is a vexing problem because it involves determining whether completed, legally effective transactions should nevertheless be denied effect for tax purposes. To say that taxpayers who do not properly implement their transactions should not receive their desired tax results does nothing to assist in determining when taxpayers who do implement their transactions properly should nevertheless be denied sought-after tax results. For these reasons, I suggest that the ineffective transaction doctrine is not a meaningful anti-avoidance doctrine at all. Two recent development suggest that the ineffective transaction doctrine is no longer even practically useful as an anti-avoidance doctrine in Canadian tax law. The first is that, after the Supreme Court of Canada's decisions in Stubart and Continental Bank Leasing Corp. v. The Queen,143 one can no longer predict with any certainty whether technical deficiencies w i l l render transactions legally ineffective. Stubart and Continental Bank suggest that minor deficiencies that do not render transactions void ab initio w i l l be treated as insignificant and wi l l not render them ineffective for tax purposes. Rather than adopting a 1 4 2 In addition, as I noted in discussing the sham doctrine, tax avoidance transactions are typically designed by, and implemented with the assistance of, professional advisors. Therefore, it is highly unlikely that mistakes will be made in carrying out the legal steps involved in those transactions. 1 4 3 98 D.T.C. 6505 (S.C.C.) [Continental Bank]. 62 strict approach to determining the legal effectiveness of transactions as a means of protecting against tax avoidance, Canadian courts seem will ing to overlook technical deficiencies in even tax-motivated transactions. The second development is the relatively new application of the equitable doctrine of rectification to tax cases. Recent cases suggest that rectification is being used to "correct" transactions that result in unanticipated tax consequences. Rectification could be used to correct transactions that would otherwise be legally ineffective. i . The Doctrine After Stubart and Continental Bank In Stubart, a holding corporation caused a profitable subsidiary, Stubart Investments Ltd. ("Stubart"), to sell its assets to a sister subsidiary, Grover Cast Stone Co. ("Grover"), which had accumulated tax losses. After acquiring legal ownership of Stubart's assets, Grover appointed Stubart as its agent to carry on the business on Grover's account. Stubart continued to operate the business as it always had, but at the end of its year paid its profit to Grover, which then offset those profits with its accumulated tax losses. The transfer was motivated solely by tax purposes, and was an attempt to use Graver's tax losses to shelter a related corporation's income. The transaction effectively consolidated profits and losses within a related group of corporations, which is not normally permitted by the Act. One of the Crown's arguments for treating the income as Stubart's was that the asset transfer was legally ineffective because it was incomplete in a number of respects. 1 4 4 The Supreme Court 1 4 4 The Crown identified the following deficiencies in the transaction: no part of the purchase price was allocated to goodwill, Stubart failed to transfer licenses required under the Excise Tax Act, corporation information returns were not filed as required by provincial corporate law, Stubart continued to show its name on the business premises and to pay for utilities and property taxes, Stubart filed T-4 slips for the employees' of Canada was not convinced that the Crown had identified any fatal failures in the transaction and concluded that the asset transfer was effective despite the failure to observe all technicalities. 1 4 5 In Continental Bank, Continental Bank (the "Bank") decided to sell its leasing subsidiary to a third party, Central Capital Leasing ("Central"). For commercial reasons, Central preferred to purchase the subsidiary's assets rather than its shares. However, for tax reasons, the Bank preferred to sell the subsidiary's shares. The Bank therefore caused the subsidiary to form a partnership with Central's subsidiaries and to roll its assets into the new partnership on a tax-deferred basis under subsection 97(2) of the Act. The subsidiary was wound up three days later and its assets, including its interest in the new partnership, were distributed to the Bank . 1 4 6 The sole purpose for creating the partnership structure and for carrying out the tax-deferred rollovers was to enable the Bank to subsequently sell the partnership interest (representing a 99% entitlement to the leasing subsidiary's assets) to Central (who already controlled the other 1%) at a tax cost roughly equal to the cost of a share transaction. The tax-deferred rollover into the partnership was the key step in this series of transactions because that rollover fixed the cost of the partnership interest that the Bank ultimately sold to Central. The Crown argued that the asset rollover was ineffective on the basis that the partnership itself was invalid for two reasons. First, the Crown argued that the partnership was tax-motivated rather than profit-motivated, and therefore did qualify as a "partnership" earnings, and no notice of change in ownership was sent to creditors, customers, suppliers, and others. The Federal Court of Appeal, referring to its duty under Atinco Paper Products, concluded that the transaction was not a "complete, real transaction." See 81 D.T.C. 5120. 1 4 5 Among the facts that Estey J. relied on is that Grover subsequently sold the business assets to an unrelated third party. Estey J. observed that it is difficult to reconcile this commercial reality with a finding that the transfer from Stubart to Grover was legally ineffective. It is not clear from the Court's reasons what the Crown had argued in response to this point. 1 4 6 The distribution was done using a tax-deferred rollover under subsection 88(1) of the Act. 64 under the Ontario Partnership Act.147 Second, the Crown argued that the partnership was not 148 legally valid because the federal Bank Act prohibits a bank from participating, directly or indirectly, in a partnership, and because the Ontario Partnership Actm provides that a partnership is dissolved if it is unlawful for one of its members to be a partner. In relying on the Bank and Partnership Acts, the Crown effectively invoked the ineffective transaction doctrine. 1 5 0 A s in Stubart, the Supreme Court of Canada declined to apply the doctrine. The majority concluded that the partnership was valid because the Bank's subsidiary, not the Bank itself, was legally the partner that rolled the assets into the partnership. The Court declined to look behind the subsidiary by treating the Bank as a participant in the partnership, even though the Bank controlled its subsidiary. Therefore, the partnership was validly in existence at the time of the asset rollover because there was no bar to the participation of the entity who was legally the partner. 1 5 1 After Stubart and Continental Bank, it is clear that although the ineffective transaction doctrine might apply in cases where transactions are fundamentally deficient (such as where contracts do not provide for consideration, or ownership of property does not pass), ancillary deficiencies in transactions and doubts about the legal validity of a taxpayers' acts are insufficient to render those transactions ineffective. Based on Stubart, Arnold and Wilson suggest that only substantially-completed transactions likely wi l l not be ignored for 1 4 7 Section 2 of the Ontario Partnership Act, R.S.O. 1980, c. 370, defines "partnership" as "the relation subsisting between persons carrying on business in common with a view to profit." The Crown argued that there was no profit motive in forming the partnership, merely a tax motive, and that therefore a partnership had not come into being. The Supreme Court unanimously held that, despite the predominant tax motive, the partners did have an ancillary profit-making and profit-sharing motive, and therefore rejected the Crown's argument on this point. 148 Bank Act, R.S.C. 1985, c. B-l , para. 174(2)(i). 149 Partnership Act, supra note 147, s. 34. 1 5 0 The Crown succeeded at the Federal Court of Appeal, which held that the partnership was void ab initio due to the Bank Act and the Partnership Act. The Federal Court of Appeal held that the real transaction was a sale of assets by the subsidiary directly to Central. See 96 D.T.C. 6355. 1 5 1 Two Supreme Court Justices dissented on this point, concluding that the partnership was void because of the combined operation of the Bank Act and Partnership Act. tax purposes. Similarly, Continental Bank suggests that Canadian courts w i l l apply the ineffective transaction doctrine strictly. It therefore appears that the doctrine, like the sham transaction doctrine, has a relatively narrow scope in Canadian tax law. i i . Rectification It is also possible that, even if the ineffective transaction doctrine were truly an anti-avoidance doctrine, its usefulness as a tool to control tax avoidance is at an end. In several recent cases, taxpayers faced with unexpected tax consequences have sought to rectify their transactions, and they have met with some success. For example, in Canada (Attorney General) v. Juliar,153 taxpayers attempted to execute a tax-deferred share rollover of shares in a holding corporation but erroneously took back promissory notes, instead of shares, as consideration for their shares. The result was a deemed dividend under section 84.1 of the Act, not a tax-deferred rollover under section 85. The taxpayers convinced the Ontario Court of Appeal to rectify the transaction and to substitute shares of the transferee corporation for the promissory notes originally issued. The Court did so because it was satisfied that the taxpayers had always intended the transaction to be a tax-deferred rollover, and that the transaction would have qualified as one had the taxpayers executed it properly. Rectification is an equitable remedy available when, due to a mistake or inadvertence, a transaction has not been legally completed or the documents entered into do not properly reflect the terms actually agreed upon by the parties. Rectification is not limited to the mechanics of a transaction, but may also extend to the fundamental elements of the 1 5 2 Arnold & Wilson, "Part I", supra note 74 at 853. 1 5 3 2000 D.T.C. 6589 (Ont. C.A.), affirming 46 O.R. (3d) 104 (Ont. Sup. Ct.). 66 transaction itself, such as the identity of a party to a contract. 1 5 4 Because rectification permits taxpayers to "cure" technical defects in transactions that have resulted in unanticipated tax consequences, it is possible that the advent of rectification signals the end of the ineffective transaction doctrine. Although the Canadian tax authorities w i l l always be entitled to challenge transactions that do not validly establish the requisite legal relationships, such an argument would be practically useless if taxpayers were able to have their transactions rectified by other courts, whose orders are binding on the tax authorities and tax courts. 1 5 5 To the extent that taxpayers obtain rectification for the purpose of changing the tax consequences of their transactions, either by correcting those transactions or by substituting other transactions with more favourable tax consequences, rectification could facilitate tax avoidance by permitting retroactive tax planning. To date, there has not been a case where a taxpayer sought to rectify an explicit tax avoidance scheme that had not been fully implemented. Therefore, it is impossible to predict how significant the tax motivation would be to a court considering rectification, 1 5 6 or whether 1 5 4 In Snow White Productions Inc. v. PMP Entertainment Inc. et al., 2005 D.T.C. 5150 (B.C.S.C.) [Snow White], the Court rectified a television production services agreement and changed the identity of the party owning the copyright in a television movie in order to permit Snow White to qualify as an "eligible production corporation" under the Act and therefore to be eligible for federal tax credits. The Court was satisfied that the parties to the agreement had always shared a common intention to obtain the tax credits in question. However, in 771225 Ontario Inc. v. Bramco Holdings Ltd. (1995), 21 O.R. (3d) 739 (Ont. C.A.), affirming 17 O.R. (3d) 571 (Ont. Gen. Div.) [Bramco], the Court denied an application to rectify a transfer of land. The transferor had mistakenly identified the wrong corporation as the transferee, giving rise to unexpected and unintended provincial land transfer tax consequences. At 742, the Ontario Court of Appeal invoked a "well-established rule in tax cases that the courts do not look with favour upon attempts to rewrite history in order to obtain more favourable tax treatment." 155 Dale v. R., 97 D.T.C. 5252 (F.C.A.) establishes that orders of other courts on private law matters, even retroactive orders, are binding on Canadian tax authorities and tax courts. 1 5 6 In Juliar, supra note 153, the taxpayer's underlying intention was to execute a tax-deferred rollover. Such rollovers under section 85 of the Act are commonplace business and tax planning transactions, and nothing about the Juliars' request for rectification smacked of abusive tax avoidance. Drawing the line between genuine rectification and retroactive tax planning will hinge on the facts of each case. If taxpayers facing unexpected tax consequences are able to produce convincing evidence of legitimate, bona fide non-tax purposes, the courts will likely permit them to rectify transactions in order to attract the originally-intended tax consequences. 67 the rule against rewriting history cited in Bramco151 w i l l continue to apply. It is possible that a court would decline to rectify a transaction that originally had no bona fide purpose other than tax avoidance, but the law in this area remains unclear. In any event, Juliar and Snow White are precedents for rectifying transactions based on a general intention to mitigate tax, and that alone might mean the end of the ineffective transaction doctrine in Canadian tax law. D. The Business Purpose Test The business purpose test is a third anti-avoidance doctrine. Generally, the business purpose test holds that a tax-reducing transaction wi l l be given effect for tax purposes only i f the taxpayer entered into the transaction for a legitimate business purpose. 1 5 8 The rationale for this doctrine is that tax-relieving legislative provisions should apply only to transactions with purposes that Parliament could have contemplated when deciding to grant favourable tax treatment. 1 5 9 Since Parliament would not have intended taxpayers to undertake transactions solely for the purpose of reducing tax, tax-motivated transactions without some other non-tax purpose should not be eligible for tax benefits. Explaining this reasoning another way, Gideon and Kent explain that the business purpose test is aimed at transactions that secure tax reductions in ways that " w i l l so undermine the intended tax structure that [they] cannot fairly have been within : legislative contemplation." 1 6 0 When transactions fail Supra note 154 at 742. Arnold, "Business Purpose Test", supra note 100 at 10:11. Tiley, "Part II", supra note 100 at 221. Gideon & Kent, supra note 35 at 7:13. 68 the business purpose test, taxpayers' tax consequences wi l l be determined without giving those transactions any effect for tax purposes. 1 6 1 The business purpose test originated in the decisions of the United States Court of Appeals and Supreme Court in Gregory v. Helvering.162 In that case, the taxpayer, Gregory, wished to transfer shares of one corporation ("Monitor") held through a holding company ("United") from United to herself. To avoid the dividend that would have resulted had United simply transferred the Monitor shares directly to her, Gregory had United transfer the shares to a new corporation ("Averill") in exchange for Aver i l l shares issued directly to Gregory. Aver i l l was subsequently wound up and its assets, being the Monitor shares, were distributed to its shareholder Gregory. Under the U S tax code at the time, the transfer of shares to Aver i l l and the subsequent distribution of those shares to Gregory on Aver i l l ' s winding up would have been tax-exempt if undertaken " in pursuance of a plan of reorganization." 1 6 3 The Court concluded that, because Aver i l l served no business purpose and was incorporated solely to facilitate the circuitous transfer of shares to Gregory, the transfers had not been made " in pursuance of a plan of reorganization." The Court held that the only reorganizations that the legislature would have intended to benefit from the exemption were reorganizations undertaken in the course of business of a corporation. 1 6 4 The Court therefore declined to treat the Aver i l l transactions as the type of reorganization contemplated by the provision. The business purpose test expressed in Gregory v. Helvering is strictly confined to transactions undertaken in a business context. Therefore, transactions undertaken outside a 1 6 1 Arnold, "The Business Purpose Test", supra note 100 at 10:11. 162 Supra note 101. 1 6 3 Under section 112(g) of the Revenue Act of 1928, c. 852, 45 Stat. 791. 164 Gregory v. Helvering, supra note 101 at 469. 69 business context would not be caught by this doctrine, even i f arranged for no purpose other than tax reduction. In my view, the restricted scope of that test is inadequate in a complex modern economy, in which many transactions that are not business transactions have tax consequences and may present opportunities for tax avoidance. 1 6 5 To address this concern, the business purpose test has evolved into a broader non-tax purpose test applicable in all contexts. The non-tax purpose test requires tax-reducing transactions to be justified by some reason other than tax reduction. 1 6 6 Under this formulation of the test, any economic rationale, even important personal objectives such as estate planning, justifies transactions with tax-reducing effects. 1 6 7 A s Arnold argues, the evolution of the business purpose test into a broader non-tax purpose test is reasonable given the wide range of non-business transactions that nevertheless have tax consequences and would normally lack a strict business purpose. 1 6 8 It would be unreasonable and inequitable to apply a business purpose test only to business transactions and not to others. It would be equally unreasonable to deny beneficial tax consequences to transactions outside the business context merely because those transactions did not serve a business purpose. In this regard, the general non-tax purpose formulation of the business purpose doctrine makes for a more equitable tax system. Furthermore, where Parliament uses tax incentives to encourage taxpayers to engage in certain non-business activities (such as saving for retirement), denying the benefit of those incentives because transactions 1 6 5 Such transactions designed to stream income to a non-taxable spouse or minor child, or to make charitable donations of property acquired with borrowed funds in order to secure tax refunds that exceed the actual cash outlay. 1 6 6 Gideon & Kent, supra note 35 at 7:4. Also Arnold, "Business Purpose Test", supra note 100 at 10:11. 1 6 7 Gideon & Kent, ibid, at 7:5. The more general non-tax purpose formulation originated in the U.S. Supreme Court judgment in Knetsch v. United States, 364 U.S. 361 (1960). Tiley observes that, in Gregory v. Helvering, the Court had rejected such a broad doctrine, only to endorse it 25 years later in Knetsch. Tiley, "Part II", supra note 100 at 223. 1 6 8 Arnold, "Business Purpose Test," supra note 100 at 10:12. 70 undertaken to obtain them do not have a business purpose would deter taxpayers and defeat Parliament's objectives. 1 6 9 Finally, the broadened non-tax purpose test also recognizes that tax avoidance is not confined to business contexts but may equally occur in other contexts. The broad version of the business purpose test remains viable in American tax law. However, there is no business purpose test in either U K or Canadian tax law. English law generally permits taxpayers to arrange their affairs in order to reduce tax. In English law, a tax purpose is not a basis for disregarding the effect of a transaction - it is simply not relevant to the tax consequences unless the statute explicitly makes the purpose relevant. English law generally adheres to the main principles enunciated in Duke of Westminster,™ 171 which are fundamentally inconsistent with a business purpose requirement. A s in the U K , there is no business purpose test in Canadian tax law. Indeed, the Supreme Court of Canada definitively rejected one in Stubart}12 In Stubart, the Court rejected a business purpose test for two reasons. First, the Court believed that a business purpose test is inconsistent with a complex revenue statute that includes many tax incentive provisions. The Court suggested that disregarding transactions 1 6 9 Tiley, supra note 100 at 222. Also Gideon & Kent, supra note 35 at 7:3-4. This point was also raised in Stubart, supra note 31 at 6322, as a reason for rejecting outright a general business purpose test in Canadian tax law. Unfortunately, the Court referred to an unduly narrow formulation of the doctrine, and therefore missed the point with respect to the utility of a business purpose test in achieving the underling policy, or object and spirit, of the Act, which the Court said should be the goal of statutory interpretation. 1 0 Supra note 53. 1 7 1 Arnold argues that the business purpose test is not actually inconsistent with the Duke of Westminster principles because the business purpose test simply ensures that taxpayers are truly arranging their affairs rather than just manipulating tax consequences. "Business Purpose", supra note 100 at 10:21. I am not convinced by this argument, however, because in speaking of taxpayers truly arranging their affairs Arnold seems to be implying a requirement that transactions have real economic substance. Canadian tax law is clear that tax consequences are a function of the legal relationships created by transactions. There is no requirement that transactions have real economic substance as long as they are legally effective and they create real legal relationships. Therefore, taxpayers may be able to rearrange their affairs by rearranging the legal form of their affairs, without any change to their underlying economic substance. 1 7 2 Krishna, supra note 2 at 1004. The Federal Court of Appeal had once recognized a form of business purpose test in Leon, supra note 132, discussed above under the heading "The Sham Transaction Doctrine." However, that case appears to have confused the business purpose test with the shame doctrine. As discussed above, the courts quickly retreated from Leon and it was expressly rejected in Stubart. 71 arranged solely to obtain those tax incentives because they have no business purpose would 173 inhibit taxpayers from undertaking the very activities that Parliament seeks to encourage. In addition, the Court concluded that a judicial general anti-avoidance doctrine was not necessary in Canada because the Act already contained a form of general anti-avoidance provision (the former section 137, later renumbered subsection 245(1) prior to the introduction of the G A A R ) . 1 7 4 In the Court's view, that provision constituted a legislative response to "blatant practices designed to defeat the Revenue," 1 7 5 and therefore judicial anti-avoidance doctrines should be developed only in future cases under that provision. Stubart did not, however, completely reject a business purpose test in all instances. In his reasons, Estey J. formulated a series of "interpretive guidelines" intended to assist courts dealing with tax avoidance. In addition to suggesting that section 137 itself might be interpreted in the future as including a form of business purpose test, Estey J. proposed that: .. .the business purpose doctrine is an appropriate tool for testing the tax effectiveness of a transaction, where the language, nature and purposes of the provision.. .indicate a function, pattern and design characteristic solely of business transactions. 1 7 6 Despite the fact that Estey J.'s interpretive guidelines left some room for a judicial business purpose test in Canadian tax law, if only in limited circumstances, subsequent cases have not 1 3 Stubart, supra note 31 at 6322. 1 7 4 Arnold & Wilson, "Part I", supra note 74 at 837-38. Arnold & Wilson argue that section 137, rather than concern about tax incentives, was the Court's main reason for rejecting the business purpose test completely. 175 Stubart, supra note 31 at 6321. 176 Ibid, at 6324. I believe that this statement, combined with the argument that requiring a business purpose for all tax-reducing transactions would defeat Parliament's purposes by discouraging taxpayers from engaging in transactions for the purpose of obtaining tax incentives, displays a fundamental misunderstanding of the business purpose test. Estey J. clearly limited his analysis to the "narrow" business purpose test applicable only to business functions, and failed to appreciate that the test can be more broadly applied as a non-tax purpose test. However, Canadian cases since Stubart have reaffirmed that there is no business purpose test in Canadian tax law. Furthermore, the introduction of the GAAR, which includes a form of non-tax purpose test, renders any debate of Estey J.'s reasoning moot. 72 applied or developed those guidelines in a meaningful way. No business purpose test has 177 emerged in relation to any provisions of the Act that do not expressly include one. E. The Step Transaction Doctrine The step transaction doctrine is a fourth judicial anti-avoidance doctrine, applied particularly in the United States and the U K . This doctrine focuses on series of inter-related transactions undertaken for legitimate ultimate purposes, into which are inserted legally effective intermediate transactions (or steps) with no purpose other than reducing the tax payable on the outcome of the series. Generally, for tax purposes, each transaction in a series of transactions must be examined independently and be given its separate legal effect. Therefore, in determining the tax effects of a series of transactions, normally each transaction would be analyzed in a step by step manner and would give rise to its own distinct tax consequences. The step transaction doctrine, by contrast, permits a court to disregard the steps inserted into a series merely for tax avoidance purposes, and to tax the end result of the series without regard to the tax consequences of the impugned step. 180 For instance, in Furniss (Inspector of Taxes) v. Dawson, a taxpayer wishing to sell shares in one corporation to a third party first exchanged them for shares in a new 1 7 7 Such as paragraph 18( 1 )(a) of the Act, which imposes a business purpose test for business expense deductions. > 1 7 8 Arnold & Wilson, "Part I", supra note 74 at 859. This doctrine could possibly be viewed merely as an application of the business purpose test to the intermediate steps in a series of transactions and not as a separate anti-avoidance doctrine. However, since the doctrine involves a departure from the general requirement to give effect to each transaction in a series in a step-by-step fashion, I prefer to characterize the step transaction doctrine as a distinct anti-avoidance doctrine, while recognizing that it is related to the business purpose test. 179 Ibid, at 859. In the UK, a series of transactions with inserted tax avoidance steps is referred to as a "composite" transaction. In W.T. Ramsay Ltd. v. I.R.C., [1982] A.C. 300 (H.L.) [Ramsay], the leading UK case adopting a step transaction doctrine, Lord Wilberforce, at 323, dismissed the general step-by-step rule mentioned above as inappropriately compelling the court to look at each transaction in a series in "blinkers," isolated from the overall context in which it belongs. 1 8 0 [1984] A.C. 474 (H.L.) [Furniss]. 73 corporation on the understanding that the new corporation would then sell the shares to the purchaser. The share exchange deferred the realization of any capital gains in the taxpayer's hands until the sale of the new corporation's shares. The inserted step, the share exchange, served no purpose other than to avoid tax otherwise payable had the shares been sold directly to the purchaser. Similarly, in Canadian Pacific,m the taxpayer needed Canadian funds to use in its business, but first borrowed Australian funds and then converted them into Canadian funds because the prevailing Australian interest rate was higher. 1 8 2 The intermediate step of borrowing of Australian funds and converting them into Canadian resulted in an interest deduction at the higher Australian rate instead of the lower Canadian rate. However, that step served no business purpose since Canadian Pacific needed Canadian funds - it was undertaken solely because it resulted in lower taxes than borrowing Canadian funds directly. A court applying the step transaction doctrine to the Furniss and Canadian Pacific transactions would disregard the tax consequences of the intermediate steps and would instead impose tax based on the end result of the overall series. Thus, the court would treat the taxpayer in Furniss as if he had disposed of his shares directly to the purchaser, and Canadian Pacific as if it had borrowed Canadian funds directly. The step transaction doctrine thereby permits a court to disregard the tax consequences normally resulting from legally 181 Supra note 99. 1 8 2 The Canadian Pacific transaction is colloquially known as "weak currency borrowing.". In addition to first borrowing Australian funds, Canadian Pacific entered into a forward exchange contract for a future purchase of Australian funds in order to repay the loan when due. At the time of the transaction, the Australian dollar was gradually declining in value as compared to the Canadian dollar, so that in the future it would require fewer Canadian dollars to repay the loan in Australian funds than the number of Canadian dollars originally yielded on conversion of the original Australian principal. Therefore, by locking in at a lower projected exchange rate, Canadian Pacific guaranteed itself a foreign exchange gain resulting from the difference between the Canadian dollars received initially and the Canadian dollars later required to repay the loan. That gain partially offset the higher interest costs resulting from borrowing Australian funds because the gain was taxed at the lower capital gains inclusion rate while 100% of the interest was deducted. A similar transaction occurred in Shell, supra note 99. 74 effective intervening steps and to impose tax based on the ultimate result achieved by a taxpayer. Under this doctrine, therefore, a court is not required to give effect to the tax consequences of each step in a series in turn before considering the next step. 1 8 3 The theory underlying the step transaction doctrine is that, despite the presence of tax-reducing intermediate steps, a series of interrelated transactions is more appropriately treated as one transaction when the intermediate steps have no purpose other than to secure certain tax consequences for the series as a whole. This proposition is particularly apposite when the inserted steps are self-cancelling transactions without ongoing legal consequences, such as transitory occurrences or "detours" taken by taxpayers in the course of arriving at particular outcomes. Such transitory "detours" occurred in Ramsay,184 where a taxpayer with capital gains undertook a prearranged transaction designed to create offsetting gains and losses, of which the gains were non-taxable and the losses could be used to shelter the 1 8S original gain; and in Gregory v. Helvering, where the taxpayer incorporated Aver i l l for the sole purpose of acquiring and distributing the Monitor shares, after which it was dissolved. In both cases, the inserted steps were self-cancelling in that their effects were subsequently undone by subsequent transactions in the series. 1 8 6 When intermediate steps do not themselves have an enduring non-tax legal effect, it is relatively easy to justify applying this doctrine on the basis that the transactions are merely I 87 "empty formalities" and should be ignored. However, the doctrine also applies to transactions that have real legal consequences but which are so closely inter-related with the 1 8 3 Tiley, "Part II", supra note 100 at 235. 184 Supra note 179. 185 Supra note 101. 1 8 6 Another transitory "detour" occurred in I.R.C. v. Burmah Oil Co. Ltd., [1982] S.T.C. 30 (H.L.) [Burmah Oil], in which the taxpayer entered into a circular transaction using self-cancelling loans to convert a non-deductible debt into an allowable capital loss. 1 8 7 Tiley, "Part II", supra note 100 at 236. overall series that they should nonetheless be "integrated" into a larger single transaction. 1 8 8 To develop a rationally justifiable doctrine that distinguishes between intermediate transactions whose legal effects wi l l be ignored or integrated into larger transactions, and intermediate transactions whose legal effects wi l l be respected, requires identifying a degree of connection, or nexus, between the intermediate step and the overall series that wi l l justify applying the doctrine. Tiley concludes that courts have formulated the relevant nexus in three different ways . 1 8 9 Each formulation involves a different degree of connection between the tax avoidance steps and the remainder of the series. The first nexus is a "binding commitment," once the tax avoidance step is taken, to complete the subsequent steps in the series. 1 9 0 The binding commitment test suggests that intermediate steps may be disregarded for tax purposes only when the nexus between those steps and the remainder of the series is so close that completion of the series after those steps are taken is pre-ordained. This version of the nexus was applied in Ramsay, 1 9 1 where it was pre-ordained that the taxpayer would see the entire series through upon undertaking the tax avoidance step. Similarly, in Furniss the House of Lords found that it was pre-ordained that the intermediate corporation would dispose of the shares to the third party purchaser to whom the taxpayer originally intended to sell. Therefore, their Lordships were prepared to disregard the intermediate share transfer and to consider the series as one transaction constituting a sale to the third party purchaser. 188 Ibid, at 236. 189 Ibid, at 235. 190 Ibid, at 235. The United States Supreme Court endorsed the binding commitment test in Commissioner of Internal Revenue v. Gordon, 391 U.S. 83, 96 (1968). 191 Supra note 179. '92 Supra note 180. 76 The second nexus is a more relaxed connection, present when the steps in a series are "so interdependent that the legal relations created by one transaction would [be] fruitless" without completing the series. 1 9 3 Such a nexus would exist in a weak currency borrowing case like Canadian Pacific.194 In that case, the conversion of Australian funds into Canadian may not have been pre-ordained because Canadian Pacific had no legal obligation (such as a contractual commitment) to do so, but the Australian dollar borrowing would have been fruitless without the subsequent conversion into Canadian funds. Under the "mutual interdependence" test, therefore, the step transaction doctrine would apply and Canadian Pacific would be treated as if it had borrowed Canadian funds directly. The intermediate Australian dollar borrowing would be ignored. On the other hand, this nexus might not exist in the Furniss transaction. 1 9 5 The incorporation of an intermediate corporation to hold the shares would not necessarily be fruitless without a subsequent sale to a purchaser, because there could be non-tax reasons for that transaction. For the same reason, this nexus might not exist in a case like Gregory v. Helvering - although in that case the evidence revealed that the dissolution of Aver i l l was pre-ordained, thus satisfying the more demanding binding commitment test. The third nexus is even more relaxed and is known as the "end result" test. This nexus exists where the intermediate steps are really components of a single transaction and were intended from the outset to reach an ultimate result, even i f it is not certain at the outset 1 9 3 Tiley, "Part II", supra note 100 at 235. Also Redding v. Commissioner of Internal Revenue, 630 F.2d 1169, 1177 (7* Cir. 1980). In Ramsay, supra note 179, Lord Wilberforce suggested that, in addition to cases of a legally binding commitment to complete series, the doctrine would apply where there is "no practical likelihood" that series would not be completed after the tax avoidance step. It is not clear whether Lord Wilberforce meant to describe a test equivalent to a legally binding commitment, or whether he intended to establish a secondary, slightly less restrictive threshold for the doctrine along the same lines as "mutual interdependence." 194 Supra note 99. l9SIbid. at 235. 77 whether the series w i l l be completed. This nexus was employed in McDonald's Restaurants, in which the taxpayers purchased restaurants for stock in March and registered the stock in September so that the vendors could sell it for cash. The taxpayers argued for treating the restaurant sale and the sale of the stock for cash as a single transaction because it had favourable effects on the basis of the restaurant assets for depreciation purposes. The Court of Appeals concluded that the "end result" test was satisfied and amalgamated the transactions on the basis that "they were really component parts of a single transaction intended from the outset to be undertaken for the purpose of reaching the ultimate result." Courts in the United States have been wil l ing to apply all three versions of the step transaction nexus. 1 9 8 In the U K , on the other hand, the courts have been reluctant to expand the doctrine beyond the binding commitment test. 1 9 9 U K courts have consistently required a significant degree of pre-ordination before departing from the general rule of giving effect to each step in a series of transactions in a step-by-step fashion. For instance, in Craven v. White,200 the House of Lords distinguished a situation in which taxpayers intending to sell shares transferred the shares to an offshore holding company from Furniss on the basis that the taxpayers abandoned their negotiations with the purchaser and did not complete a sale. 196 Ibid, at 235-36, citing McDonald's Restaurants of Illinois Inc. v. Commissioner of Internal Revenue, 76 T.C. 972, 994 (1981) [McDonald's Restaurants]. 197 McDonald's Restaurants, ibid, at 524. Ironically, the US Tax Court had identified this version of the step transaction doctrine but had declined to apply it. The Tax Court instead applied the binding commitment version and concluded that the test was not satisfied. 1 9 8 Gideon & Kent, supra note 35 at 7:9, observe that the binding commitment, mutual interdependence, and end result versions of the step transaction doctrine continue to coexist in United States case law and are not always consistent. They suggest that the length of time separating the steps in the series of transactions is a key factor in determining the degree of preordination required before distinct steps will be integrated into one transaction. Long Term Capital Holdings v. United States, 330 F. Supp. 2d 122, 191 (D. Conn., 2004) [Long Term Capital Holdings], a recent U.S. case, acknowledges that all three branches of the step transaction doctrine remain viable. 1 9 9 The UK courts resisted any form of the step transaction doctrine for a long time after American courts had begun to apply it. UK law remained wedded to a literal, step-by-step approach until the early 1980s, when the House of Lords released its revolutionary judgments in Ramsay, Burmah Oil, and Furniss in succession. 2 0 0 [1989] A.C. 398 (H.L.). 78 Despite a continuing intention to ultimately sell their shares, to whom and when the sale would take place were unknown and not pre-ordained. The majority of law Lords would have integrated the transactions only if the scheme were planned as a whole and there was no practical likelihood that the pre-planned scheme would not be carried out. "No practical likelihood" meant that the intermediate step was not contemplated as having an independent life separate from the series as a whole. This threshold is clearly higher than the American "end result" test applied in McDonald's Restaurants. 201 In a series of recent cases, including I.R.C. v. McGuckian, MacNiven v. Westmoreland Investments Ltd.,202 and Barclays 2 m the House of Lords has retreated further from the step transaction doctrine expressed in Ramsay and Furniss. In these cases, the House of Lords has rejected treating the doctrine as a distinct anti-avoidance doctrine and has re-emphasized "the primacy of the statutory language." 2 0 4 These cases, particularly Barclays, have effectively "killed off the.. .doctrine as a special theory of revenue law and 90S subsumed it within the general theory of statutory interpretation." The House of Lords' retreat has been motivated primarily by concerns that the threat of the doctrine created "an unacceptable level of unpredictability which stifles invention on the one hand and commercial planning on the other." 2 0 6 In addition, their Lordships have expressed a desire to reconcile the doctrine with the modern approach to statutory interpretation, which requires courts interpreting tax statutes, like any others, to give effect to Parliamentary intent. Their Lordships have therefore emphasized that the step transaction doctrine is not a general anti-2 0 1 [1997] S.T.C. 908 (H.L.). 2 0 2 [2003] 1 A.C. 311 (H.L.). 203 Supra note 80. 2 0 4 Leonard Hoffmann, "Tax Avoidance" [2005], no. 2 British Tax Review 197 at 202. 205 Ibid, at 203. 206 Ibid, at 204. 79 avoidance rule existing outside the statute, but is in fact merely a matter of statutory interpretation. 2 0 7 In doing so, they have effectively re-characterized Ramsay as a sort of awakening from the tradition of literal interpretation and an adoption of a more modern approach to interpretation. The message of these recent cases is that tax avoidance should be dealt with primarily through statutory interpretation rather than through special anti-avoidance doctrines, so that tax avoidance transactions not contemplated by the statute are struck down on the basis that Parliament did not intend those transactions to qualify for tax benefits. 2 0 8 Unlike in the United States and the U K , there is no step transaction doctrine in Canadian law, even in the face of blatant tax avoidance. Canadian courts have continued to approach series of transactions in a step by step fashion, treating each transaction as distinct and giving them effect for tax purposes regardless of their purpose. I believe that, because the step transaction doctrine is intimately related to the business purpose test in that it essentially applies a business purpose test to the individual components in a series, Canadian courts have been strongly influenced by the wholesale rejection of a business purpose test in Stubart. Although there may have been rare cases in which Canadian courts have ignored the effects of intermediate transactions, 2 0 9 there have been no such cases since Stubart. Furthermore, when the Supreme Court of Canada considered weak currency borrowing in 207 Ibid, at 202-03. In "The Control of Avoidance: The United States Experience" [1998], no. 2 British Tax Review 161, an article published shortly after the McGuckian decision, John Tiley and Erik Jensen argue at 169 that anti-avoidance doctrines must "rest on" statutory interpretation because "that is the authority for their being part of the legal system, whether in the United States or in the United Kingdom." Oddly, in I.R.C. v. Scottish Provident Institution. [2004] U.K.H.L. 52, released the same day as Barclays, the same panel of law Lords, professing to interpret the statute in a purposive way, nevertheless invoked the Ramsay decision as a basis for treating a series of transactions as an integrated whole and denying the taxpayer a resulting tax benefit. On the basis of these two cases, the status of the doctrine in UK law remains unclear. 2 0 9 Arnold & Wilson, "Part I", supra note 74 at 859, suggest that Canadian courts have done so in some cases; however, they acknowledge that these instances have been rare. 80 Shell210 and an interest deductibility swap in Singleton,2U the Court permitted the insertion of intermediate steps without a business purpose. In fact, the Court re-emphasized the importance of respecting the legal relations created by taxpayers at each stage in a series of transactions. Absent a specific statutory provision to the contrary, therefore, it is clear that Canadian tax law does not include any form of the step transaction doctrine. 2 1 2 F. Economic Substance "Substance over form," or "economic substance," is another anti-avoidance doctrine endorsed by courts in other jurisdictions. However, the concept of "substance" does not have the same meaning in every jurisdiction. A s explained in Chapter 3, in my view "economic substance" refers to the underlying economic results of transactions, whereas "form" refers to the structure of the legal relationships through which taxpayers achieve those economic results. 2 1 3 In American law, "economic substance" means that transactions have economic benefits separate and distinct from any tax benefits such that there is a meaningful change in the economic position of the taxpayers. B y contrast, in the U K , "substance" refers to the real legal effect of taxpayers' transactions as opposed to the manner in which taxpayers have characterized them. It does not refer to taxpayers' underlying economic circumstances. 210 Supra note 99. 211 Supra note 89. In Singleton, the taxpayer, a lawyer, wished to purchase a house. Because mortgage interest is not deductible while interest on borrowed funds used to earn income is deductible, the taxpayer first withdrew funds from his equity account in his law firm to purchase the house and then borrowed and re-invested the borrowed funds in his equity account, rather than simply borrowing funds to purchase the house. The Court was unconcerned by the insertion of the intermediate step, which served no purpose other than to qualify for an interest deduction which would not have been available had the taxpayer simply purchased the house directly with borrowed money. 2 1 2 In Chapter 5,1 will argue that the GAAR, particularly paragraph (b) of the definition of "avoidance transaction" in subsection 245(3), is such a statutory provision to the contrary and creates a step transaction test in Canadian law. 2 1 3 Refer to the discussion in Chapter 3 under the heading "Common Approaches to Identifying Tax Avoidance - Form." 81 Since the Act normally imposes tax consequences on the basis of the legal relationships involved in taxpayers' transactions rather than their underlying economic outcome, taxpayers often select the legal form through which to achieve desired economic results on the basis of the tax consequences of those legal relationships. For instance, in Duke of Westminster,214 a Duke paying wages to his servants restructured those payments as annuity payments because wages were not tax-deductible but annuity payments were. 2 1 5 Similarly, in Singleton,216 a lawyer wishing to purchase a personal residence funded the purchase with funds withdrawn from his law firm's equity account rather than with a mortgage, and then borrowed funds to replenish the equity account because interest on income-earning funds is deductible but interest on funds used for personal purchases is not. 2 1 7 The Duke of Westminster and Singleton transactions were arranged to achieve particular economic outcomes (the compensation of servants and a personal purchase) through transactions structured to create legal relationships entitling the taxpayers to more favourable tax treatment than if they had achieved the same economic result in a more direct way. It would be tempting to assume, therefore, that a doctrine preferring substance over form would prevent both the Duke and Singleton from succeeding on the basis that the economic substance of their transactions was that in one case wages were paid and in the other funds were borrowed to make a personal purchase. A true substance over form 214 Supra note 53. 2 1 5 The Duke and his servants agreed that he would make annuity payments for seven years regardless whether the servants remained in his service, and that they would be entitled to claim their regular wages in addition to the annuity payments. However, the Duke and the servants also entered into a collateral agreement in which it was understood that the servants would not claim their wages in addition to the annuity payments. 216 Supra note 89. 2 1 7 Under paragraph 20(1 )(c) of the Act. The borrowed funds deposited back into the partnership account were secured by a mortgage on the house, the same as if Singleton had simply borrowed funds to purchase the house pursuant to a normal residential mortgage. 82 analysis, focusing on the economic reality of those transactions, would conclude that the Duke had merely "dressed up" non-deductible wage payments as annuities that, in substance, were really non-deductible wages. Similarly, a true substance over form analysis would conclude that Singleton had borrowed money for a personal purchase, not to earn income, since his net economic position after the transactions was the same as i f he had merely 218 purchased the house directly with borrowed funds under a normal mortgage. Therefore, the interest would not be deductible. Under a true substance over form analysis, both the Duke and Singleton would be taxed on the basis of the economic outcome of the transactions regardless of their legal effect. Although the actual transactions were legally different from direct payments of wages and a direct purchase of a residence, the transactions are economically equivalent and would therefore be taxed as such. 2 1 9 Interestingly, although both United States and U K law employ a type of substance over form doctrine, neither operates in the manner explained above. In fact, a consideration of the underlying economics of a transaction is not even relevant in the U K . In the U K , "substance" is taken to mean only legal substance, in the sense that tax consequences must be determined by reference to the true legal effect of a transaction regardless how it is characterized. 2 1 8 This analysis is based on comments made in Bronfman Trust, supra note 135. For the facts of that case, see the discussion of Bronfman Trust below under the heading "Substance Over Form in the UK and Canada." 2 1 9 This approach has received much criticism on the basis that it effectively substitutes alternative transactions with identical economic results for the legally effective transactions actually undertaken by taxpayers. Wilkie and Kerr are particularly critical of taxing on the basis of an "economic analogue." Scott Wilkie and Heather Kerr, "Common Links Among Jurisdictions: Informing the GAAR Through Comparative Analysis" in Report of the Proceedings of the Forty-Ninth Tax Conference, 1997 Conference Report (Toronto: Canadian Tax Foundation, 1998) 34:1-30 at 34:14. In my view, given that tax legislation reflects a Parliamentary intent to apply particular tax consequences to particular economic circumstances, a "true" substance over form approach is intellectually defensible, even though it marks a significant departure from the typical approach to applying tax law on the basis of legal relationships. 83 i . Economic Substance in the United States United States tax law disregards transactions lacking economic substance. The American economic substance doctrine does not substitute or recharacterize tax-reducing transactions by replacing them with different transactions with equivalent economic results for purposes of determining tax consequences. Rather, the American doctrine employs an economic substance analysis merely to determine whether tax-reducing transactions should be disregarded. Transactions are considered to lack economic substance when they do not reasonably have a "purpose, substance or utility apart from their anticipated tax consequences." 2 2 0 In this way, the American doctrine uses the underlying economics of transactions to evaluate whether they are abusive, but not as a means of determining the proper tax consequences. In this sense, the United States economic substance doctrine is more like a business purpose test than a "true" substance over form analysis. Transactions 221 that fail the economic substance test are simply denied effect for tax purposes. In determining whether transactions have economic substance, United States courts look at both the economic utility and the purpose of the transactions. Economic utility is present when transactions result in a meaningful change to taxpayers' economic positions. Acceptable purposes are present when transactions have a purpose other than tax avoidance. To carry out this analysis, United States courts have created a "two-pronged" test: an 220 Goldstein v. Commissioner of Internal Revenue, 364 F.2d 734, 740 (2d Cir. 1966). 2 2 1 Indeed, there is debate over the nature of the relationship between the economic substance doctrine and the business purpose test. Wilkie and Kerr, supra note 219 at 34:14, describe the economic substance doctrine as merely a "subset" of the business purpose test. By contrast, Tiley and Jensen, supra note 207 at 168, suggest that the business purpose test (as well as the sham, step transaction and other doctrines) are merely manifestations of an overarching American doctrine preferring substance over form. Similarly, in "The Economic Substance Doctrine and GAAR: A Critical and Comparative Perspective," written for the University of Toronto GAAR Symposium held November 18, 2005, at 33 (quoted with permission of the author), Jinyan Li contends that the American economic substance doctrine "incorporates" the other judicial anti-avoidance doctrines. 84 objective analysis of transactions' intrinsic economic value aside from their tax consequences; and a subjective analysis of taxpayers' motives and purposes in undertaking their transactions. 2 2 2 In addition, the courts w i l l consider whether the tax benefits resulting from the transactions are "unreasonable and unwarranted" in that they were not intended by the legislature. 2 2 3 The objective test of whether transactions have intrinsic economic value beyond their tax consequences is determined by evaluating their economic effect in various ways, such as whether they "[offer] a reasonable opportunity for economic profit.. .exclusive of tax benefits," 2 2 4 by reference to factors such as rate of return and transaction costs. 2 2 5 Other relevant considerations include whether transactions alter taxpayers' economic positions in "meaningful ways" and whether there are non-tax reasons or purposes to structure transactions in the fashion chosen by taxpayers. 2 2 6 2 2 2 Yoram Keinan, "The Many Faces of the Economic Substance's Two-Prong Test: Time for Reconciliation?" (2005) 1 N.Y.U.J.L. & Bus. 371 at 392-400. The two-pronged test is also discussed in Long Term Capital Holdings, supra note 198 at 171, and by Wilkie and Kerr, supra note 219 at 34:14. The two-pronged test emerged from the United States Supreme Court's decision in Frank Lyon & Co. v. United States, 435 U.S. 561 (1978) [Frank Lyon & Co.], although it is still unclear whether this test is unitary or disjunctive. Keinan observes, at 372-75, that some courts require a transaction to satisfy both prongs of the test, while other will respect a transaction that satisfies either prong. 2 2 3 David P. Hartion, "Sorting Out the Tangle of Economic Substance" (1998-1999) 52:2 Tax Lawyer 235 at 241. Hariton appears to be suggesting that tax incentives are enacted to encourage particular behaviours, and that the absence of economic substance or of a business purpose should not prevent taxpayers from taking advantage of such incentives. This point echoes the concerns raised by the Supreme Court of Canada in Stubart, supra note 31, discussed above under the heading "The Business Purpose Test in Canada." 224 Long Term Capital Holdings, supra note 198 at 172. 2 2 5 Prudent investors would not knowingly incur transactions costs that exceed a reasonable economic gain absent the tax benefits. Long Term Capital Holdings, ibid, at 172. Also ACM Partnership v. Commissioner of Internal Revenue, 73 T.C.M. 2189 (1997) (also known to US practitioners as "Colgate-Palmolive"). 2 2 6 Li, supra note 221 at 36. Transactions such as interposing an intermediary corporation for purposes of distributing shares, as in Gregory v. Helvering, and swapping equity for borrowed capital, as in Singleton, could be said to lack economic substance because they do not alter the taxpayer's economic position in a meaningful way. Joseph Isenbergh cites the Court of Appeals' decision in Gregory v. Helvering as the source for "first principles" on economic substance. Isenbergh, "Musings on Form and Substance in Taxation" (1982) 49 U. Chi. L. Rev. 859 at 866. Also Hariton, supra note 223, suggests that Gregory v. Helvering is really an application of an early economic substance doctrine. 85 The subjective test of taxpayer purposes requires that transactions be undertaken for purposes other than tax avoidance. 2 2 7 Taxpayers must establish, therefore, not only that their transactions have real economic results, but that they were entered into for non-tax purposes 228 that are "plausible in light of [taxpayers'] conduct and.. .economic situation." Although this test is similar to a classic business purpose test, it is not necessarily the same. The economic substance doctrine permits a substantial degree of tax motivation as long as 229 transactions are linked to taxpayers' ordinary business operations. Such a rule is an "implicit exception" to the subjective requirement for transactions that are closely related to business operations. 2 3 0 This exception recognizes that taxpayers carrying out business functions wi l l often have options for structuring their operations and wi l l choose structures with favourable tax consequences. Furthermore, such an exception recognizes the difficulty inherent in parsing a series of transactions into its discrete steps in order to evaluate each step's purpose. 2 3 1 Three conclusions may be drawn regarding the American economic substance doctrine. First, tax consequences wi l l be determined by more than merely the legal forms of transactions. Second, the economic substance doctrine is widely accepted in the United 227 Frank Lyon & Co., supra note 222 at 573-74 and 583-84. 228 Compaq Computer Corp. v. Commissioner of Internal Revenue, 113 T.C. 214, 224 (1999). Although this prong of the test is referred to as a "subjective" prong, I believe that it is more appropriately described as an objective test because taxpayers must establish that their motivations are plausible or reasonable. However, this prong could still loosely be referred to as subjective in that it relates to taxpayer intentions and purposes as opposed to purely objective characteristics of their transactions. 2 2 9 In UPS v. Commissioner of Internal Revenue, 245 F.3d 1014, 1019 (11th Cir. 2001), the Court held that a transaction has a business purpose "as long as it figures in a bona-fide, profit-seeking business." 2 3 0 Li, supra note 221 at 38. 2 3 1 Recall the weak currency borrowing in Shell and Canadian Pacific, both supra note 99. Both transactions occurred in the context of borrowing capital for legitimate business reasons, but the initial borrowing in a different currency was a tax-motivated step taken to reduce the overall tax effect of the borrowing. In Chapters 5 and 6,1 argue that Canadian courts have sometimes wrongly resisted finding that intervening steps serve no business purpose when the overall series serves a real purpose other than tax reduction. Finally, the exception addresses situations where tax-motivated transactions are neither unreasonable nor abusive, such as when tax incentives encourage particular behaviours and do not explicitly require a non-tax purpose. 86 States as an effective and necessary anti-avoidance doctrine. Finally, the doctrine is not a "true" economic substance over form doctrine because it does not impose tax consequences based on transactions' economic results. Rather, the doctrine is akin to a business purpose and step transaction doctrine but uses an economic analysis to determine whether transactions should respected. i i . Substance and Form in the U K and Canada As in the United States, U K tax law includes a doctrine giving substance primacy over form. However, in U K tax law "substance" refers to the legal substance of transactions, not to their underlying economic reality or to an economic analysis of their appropriateness. The "legal substance" of transactions means their true legal effect as opposed to how the parties have characterized them. Therefore, transactions that create the legal relationships that they purport to create (i.e., that are not shams) wi l l be effective for tax purposes, even i f they are tax-avoidance transactions with no reasonable non-tax purpose. Because the U K doctrine focuses on legal relationships as opposed to underlying economic realities, it is 233 234 fairly described as "legal substance over form" or simply "form over substance." 235 The U K substance doctrine derives from the decision in Duke of Westminster. In that case, the Revenue Commissioners argued that the Duke's payments retained the character of remuneration for services even though the Duke characterized them as annuity 2 3 2 Li, supra note 221 at 33. Also Hariton, supra note 223 at 241. 2 3 3 Brian A. Felesky and Sandra E. Jack, "Is There Substance to 'Substance Over Form' in Canada?" in Report of the Proceedings of the Forty-Fourth Tax Conference, 1992 Conference Report (Toronto: Canadian Tax Foundation, 1993)50:1 at 50:4-5. 2 3 4 Hogg et al., supra note 107 at 565-66. Also Arnold, "Reflections", supra note 10 at 12. 235 Supra note 53. 87 payments. Only one Law Lord accepted this argument. The majority held that the legally-binding annuity contracts were determinative because they were not a sham. Lord Tomlin explicitly rejected the notion of preferring economic substance to legal form, describing that idea as a "supposed doctrine" and a "misunderstanding." 2 3 7 Lord Tomlin described the substance of a transaction as "that which results from the legal rights and obligations of the parties ascertained upon ordinary legal principles," so long as the documents are bona fide 9^8 and not "used as a cloak to conceal a different transaction." Thus, Duke of Westminster established a rule that transactions' legal form governs as long as the legal documents are genuine. The underlying economic realities are irrelevant. Tax is to be imposed only on the basis of legal rights, not by reference to taxpayers' underlying economic circumstances. Canadian law adheres faithfully to the Duke of Westminster version of "substance." Traditionally, Canadian courts consider transactions' legal results to be their substance, 2 4 0 and have not been concerned with underlying economic considerations. This principle remains settled law despite a brief flirtation with economic substance in a series of Supreme Court of Canada cases beginning with Bronfman Trust.241 In that case, the trustees of a trust decided to make a capital allocation to the beneficiary, and to do so they borrowed funds rather than sold trust assts. The trust then claimed a deduction for interest paid on the That was Lord Atkin, ibid, at 15. However, Lord Atkin justified his conclusion by concluding that the collateral contract between the Duke and his servants was so closely tied to the annuities that it effectively transformed their legal effect into "remuneration." Lord Atkin did not dissent on the basis of an economic substance analysis. 2 3 1 Ibid, at 19. 2 3 8 In other words, the transaction must not be a sham. Ibid, at 21. Similarly, Lord Russell stated, at 25, that the substance doctrine means that, "having.. .ascertained the legal rights of the parties [the court] may disregard mere nomenclature and decide the question of taxability or non-taxability in accordance with the legal rights." 2 3 9 As I discuss in Chapter 5 under the heading "Statutory Interpretation and Tax Avoidance in Canada," Canadian courts also faithfully adhered to the other two main Duke of Westminster principles - that tax legislation should receive a strict interpretation, and that transactions with tax purposes are nevertheless effective. Of those three principles, only the rule of strict interpretation has even been modified in Canadian law, and even then not in a consistent fashion. 2 4 0 Arnold & Wilson, "Part I", supra note 74 at 858. 241 Supra note 135. 88 borrowed funds, arguing that those funds were indirectly used to earn income because they preserved the trust's income-earning capital . 2 4 2 The Court held that only interest directly used to earn income is deductible. 2 4 3 Furthermore, the Chief Justice speculated that, even i f the trust had sold assets to make the capital allocation and had then replenished those assets with borrowed funds, such a sale and repurchase would be nothing more than a "formality or sham designed to conceal the essence of the transaction, namely that money was borrowed and used to fund a capital allocation to the beneficiary." 2 4 4 Dickson C.J . commented that the law was moving away from strict adherence to legal forms and toward ascertaining transactions' "true commercial and practical nature." 2 4 5 He justified those comments on equitable grounds, suggesting that it would be inequitable for tax liabilities to depend upon taxpayers' relative sophistication at manipulating legal events to achieve an "aura of compliance" with the requirements for a deduction. 2 4 6 Dickson C.J . clearly recognized that the resulting economic circumstances would be identical regardless whether the trust borrowed money to make the allocation in order to preserve its capital, or first allocated capital and then replaced it with borrowed funds. In the end, capital would have been allocated, the trust would have a new debt, and the trust would hold the same value of income-earning capital. In addition, Dickson C.J . recognized the equity problems resulting when some taxpayers are better able to manipulate their tax consequences through their selection of legal forms. Bronfman Trust is clearly incompatible with Duke of Westminster principles, 2 4 7 since imposing tax consequences on the basis of 2 4 2 Recall that paragraph 20(1 )(c) permits a deduction for interest only when the borrowed funds are used to earn income from a business or property. 243 Bronfman Trust remains the seminal decision on the direct versus indirect use of borrowed funds. 244 Supra note 135 at 5068. 245 Ibid, at 5067-68. 246 Ibid, at 5068. 2 4 7 Arnold & Wilson, "Part I", supra note 74 at 858. Also Taylor, supra note 121 at 17:17. 89 underlying economic substance would mean that legal forms would no longer be controlling or could even be rendered irrelevant. The Bronfman Trust approach to transactions' "true economic and commercial nature" did not survive. Canadian courts retreated to an approach focused on legal forms and 948 re-emphasized that substance is found in transactions' legal, not economic effects. Legal relations remain determinative. This view was definitively reinforced in Shell249 and 9S0 Singleton. In Shell, the Supreme Court of Canada rejected any role for economic realities in the following terms: " . . .this Court has never held that the economic realities of a situation can be used to recharacterize a taxpayer's bona fide legal relationships." 2 5 1 Rather, the Court emphasized the need to respect legal relationships and to give them effect, regardless of transactions' purposes and economic effects. In Singleton, the Court considered essentially the same facts as Dickson C.J . had speculated in Bronfman Trust would be a "formality or sham designed to conceal the essence of the transaction." 2 5 2 Just as the Bronfman Trust would have been in the same economic position however it had funded the beneficiary's capital allocation, Singleton would have been in the same economic position however he had funded his personal purchase. The Court allowed Singleton to deduct the interest on the narrow basis that the direct use of the borrowed funds was to earn income from his law practice. 2 5 3 The Court believed that taxpayers are free to decide how use their own equity, 248 Continental Bank of Canada v. R., 94 D.T.C. 1858 (T.C.C.) at 1868-71. 249 Supra note 99. 250 Supra note 89. 251 Supra note 99 at 5676. Furthermore, at 5676 the Court observed that transactions may be recharacterized only if the labels attached by taxpayers do not properly reflect their "true legal effect." Therefore, taxpayers' legal forms are determinative for tax purposes as long as they are genuine (i.e., not shams) and legally effective. 2 5 2 Rather than purchase a house with borrowed funds, Singleton withdrew funds from his business to make the purchase and then replaced them with borrowed funds. 5 3 Interestingly, the Tax Court, the Federal Court of Appeal, and the Supreme Court of Canada each characterized the sequence of transactions differently. The Tax Court (96 D.T.C. 1850) had concluded that the taxpayer borrowed money first, put those funds into his law firm, and then paid out the funds to purchase a 90 previously invested in an income-earning pursuit, to fund personal consumption. 2 5 4 The fact that such a choice was tax-motivated is irrelevant. Furthermore, the Court dismissed Dickson C.J. 's comments in Bronfman Trust as mere obiter on a hypothetical situation and characterized Shell as rejecting Bronfman Trust. Therefore, after Shell it is clear that there economic substance plays no role in Canadian tax law. Bronfman Trust is now no more than a historical footnote, an aberration in an otherwise unbroken line adherence to the principle in Duke of Westminster. Taxpayers' legal relationships wi l l be disregarded only where their label does not reflect their true legal effect. Explained this way, the Canadian substance over form doctrine becomes merely an extension of the sham transaction doctrine, focusing solely on legal relationships without any requirement for deceit. It is not a stand-alone anti-avoidance doctrine. house. Therefore, it was impossible for the taxpayer to trace the borrowed funds to an income-earning use. The Tax Court also held that the "true economic purpose" for using the borrowed funds was to make a personal purchase. In deciding on the basis that the taxpayer first withdrew business equity to purchase the house and then replaced the equity with borrowed funds, the Supreme Court effectively recharacterized the trial Judge's factual findings, something that the Supreme Court regularly professes itself extremely reluctant to do. 254 Supra note 89 at 5538. In Singleton, the Court ignored the equity dimension raised in Bronfman Trust. By permitting Singleton to obtain an interest deduction, the Court effectively gave Singleton a tax subsidy for personal consumption, which would not be available to a taxpayer with equal income but no supply of income-earning equity from which to draw. Horizontal equity suggests that two taxpayers with equal incomes who make identical personal purchases should be taxed identically. However, the fact that Singleton enjoys reserves of capital changes the equity analysis and raises equity concerns about unequal tax subsidies for personal consumption. The Court may have ignored the equity aspect because of its comments in Shell, supra note 99 rejecting tax equity as a relevant consideration. Refer to Chapter 5, page 111, note 320, where this issue is discussed in the context of statutory interpretation in Canada. 91 G. The Abuse of Rights Doctrine The abuse of rights doctrine is an anti-avoidance doctrine found in many c iv i l law 9 SS jurisdictions, but not in Canada or other common-law jurisdictions. The abuse of rights doctrine is grounded in the proposition that it is reasonable to circumscribe the exercise of legal rights where that exercise amounts to an abuse in the sense that the rights are not being used for the purpose for which the law grants them, or in a manner injurious to others. Although the doctrine originated as a general legal proposition, not a specific tax law rule, it has been adopted to the tax context relatively easily in c iv i l law jurisdictions. The doctrine suggests that, although taxpayers may enter into any transactions that are legal, they abuse 9 S7 that right by entering into transactions solely to avoid or to reduce taxes. The doctrine therefore prevents taxpayers from arranging their affairs under private laws of property and 9S8 contract unless their transactions have bona fide purposes other than tax avoidance. In some jurisdictions, the doctrine operates as a relatively strict business purpose test, while in others the doctrine requires a court to evaluate the reasonableness of the legal forms chosen by taxpayers to achieve economic results. The abuse doctrine originated in France, where it is known as "abus de droit" 9 SQ (literally, "abuse of law"), as a general legal principle in the mid-nineteenth century. The French legislature subsequently incorporated the doctrine into tax law in 1925 (for indirect 2 5 5 This doctrine is known by a variety of names: in France as "abus de droit" or "abuse of law;" in Spain as "abuse of law;" in Germany as "abuse of legal constructions;" and in the Netherlands as "fraus legis" or "fraud on the law." The doctrine takes different forms in each jurisdiction, but each is based on similar common general civil law principles. For a more detailed review, see Vanistendael, supra note 52 at 147-51. Also Ward "Judicial and Legislative Approaches," supra note 113 at 8:3. 2 5 6 Ward, ibid, at 8:3. In Ward's description, an abuse occurs when rights are used in manner that amounts to spite or bad faith. Ward, "Business Purpose Test", supra note 114 at 68. 258 Ibid, at 69. 2 5 9 Vanistendael, supra note 52 at 135. 92 taxes) and 1941 (for income taxes). 2 6 0 In France, this doctrine disqualifies tax benefits 261 resulting only from transactions undertaken exclusively for tax avoidance purposes. Thus, 969 the French version of the doctrine is a strict business purpose test. However, French authorities have recently sought to amend the tax Act to broaden the abuse doctrine to include transactions for which tax avoidance is a dominant, rather than exclusive, purpose. Such a change would bring French law more closely in line with the common law business purpose test. German law incorporates the abuse doctrine into an "abuse of legal constructions" provision. This rule, statutorily entrenched in 1977, holds that tax liability resulting from transactions should be based on the legal forms that are appropriate to the transactions' economic results. 2 6 3 Under German law, an abuse exists when the legal forms chosen by taxpayers to achieve economic results are not appropriate to those economic results. 2 6 4 Particular legal forms are considered inappropriate or inadequate i f reasonable persons would 96S not choose those forms to achieve the desired economic goals. Unlike the French doctrine, the German doctrine does not focus on taxpayers' purposes in selecting legal forms for their transactions as much as on whether, objectively, the legal forms selected accord with economic outcomes. The German doctrine thus requires a degree of consistency between 260 Ibid, at 147. 261 Ibid, at 147. 262 Ibid, at 147. French courts also apply a second anti-abuse principle, known as "abnormal management act," which disqualifies the tax consequences of transactions that are contrary to a taxpayer's business interests. At 148, Vanistendael comments that the French tax authorities invoke this principle to prevent taxpayers from engaging in income-splitting with taxpayers at lower rates, even where there is some business purpose for doing so. 263 Ibid, at 149. 264 Ibid, at 149. 265 Ibid, at 149. 93 legal forms and economic purposes. However, this doctrine still permits taxpayers to choose lower-taxed legal forms for transactions when multiple forms are reasonably adequate. A third variation of the doctrine, known as "frau legis" ("fraud on the law"), exists in the Netherlands. The "frau legis" doctrine combines elements of a business purpose test with an objective analysis of the transaction. This doctrine was created by the Netherlands Supreme Court in 1926. 2 6 7 The doctrine applies when taxpayers elect to structure economic transactions in lower-taxed legal forms. Lower-taxed legal forms chosen by taxpayers wi l l be disregarded where the purpose of the tax legislation would be better achieved by 968 recharacterizing transactions in higher-taxed legal forms. The Netherlands doctrine applies only where tax minimization is the exclusive or dominant reason for selecting a lower-taxed legal fo rm. 2 6 9 However, taxpayers' non-tax reasons for selecting their particular legal structures must be objectively reasonable in the circumstances. The Netherlands doctrine thus combines a purpose test with an objective evaluation of the reasonableness of the chosen legal structure given the underlying economic circumstances. There is no abuse of rights doctrine in U K or Canadian tax l a w . 2 7 0 It is arguable that by enacting subsection 245(4) of the Act, which speaks of a misuse of statutory or regulatory 266 Ibid, at 149. Vanistendael points out that Austrian tax law goes even further, explicitly requiring that transactions' economic content be given precedence over their legal forms, not merely that the legal forms be economically reasonable. 261 Ibid, at 138. 268 Ibid, at 138. At 138, Vanistendael also observes that Swiss courts have applied a similar interpretation as the Dutch "frau legis" theory, holding that the law is abused when the legal forms of transactions are unusual and were entered into with the intent of obtaining tax benefits. 269 Ibid, at 138-39. 2 7 0 Although there is no explicit abuse of rights doctrine in American tax law either, Ward suggests that in creating business the purpose test, step transaction doctrine, and economic substance doctrine, American courts have accepted the basic philosophical premise underlying the abuse of rights doctrine, that individual rights must be balanced against the societal benefits achieved by such limitations. "Judicial and Legislative Approaches", supra note 113 at 115-16. 94 271 provisions, and of an abuse having regard to those provisions read as a whole, Parliament intended to incorporate an abuse of rights-type analysis into the G A A R . A s I w i l l demonstrate in Chapter 5, however, the Supreme Court of Canada has not endorsed such an interpretation of that provision. Rather, the Court elected to interpret the concept of abuse in the G A A R in terms of frustrating Parliamentary intent, which focuses the tax avoidance inquiry more on statutory interpretation and policy than on taxpayer purposes, although policy and purpose can never truly be separated. H. Conclusions Like Canada, many other advanced jurisdictions confront the challenge of controlling tax avoidance. In addition, many have found that simply increasing the number of specific anti-avoidance rules aimed at particular transactions is an unsatisfactory method of doing so . 2 7 2 Therefore, many jurisdictions either have enacted or have otherwise developed more general anti-avoidance doctrines providing a principled basis for combating tax avoidance. Some, like the United States and the U K , have not enacted general statutory rules, but have seen their courts develop interpretive doctrines to aid in identifying abusive tax avoidance. Others, such as some Western European states, have codified anti-avoidance rules based on the judicially-developed abuse of rights theory. Finally, some jurisdictions, such as 2 7 1 In French, the legislation refers simply to an "abus dans l'application des despoitions," as opposed to both a "misuse" and an "abuse" as apparently different concepts. 2 7 2 Increasing the specific anti-avoidance rules leads to increasingly complex tax legislation without preventing tax avoidance activity. Refer to Chapter 2 under the heading "Features of a 'Good' Tax System - Simplicity." 95 Australia, New Zealand, and Sweden, rely primarily on statutory general anti-avoidance rules as opposed to interpretive doctrines. Most of the anti-avoidance doctrines surveyed in this Chapter share certain principles in common. First, they all attempt to create a rational framework for disregarding transactions' normal legal effects in determining tax consequences. Second, they tend to fall into one of the three categories of approaches to tax avoidance identified in Chapter 3 - the approaches revolving around form, purpose and policy. Form-oriented approaches tend to cite a "disconnect" between a transaction's legal structure and its underlying economic result as a reason for disregarding the legal relations created. That disconnect is identified by evaluating the reasonableness of the transaction's legal structure in light of its economic outcome. Such an approach informs the American economic substance doctrine and some of the abuse principles in Western European c iv i l law jurisdictions. Purpose-oriented approaches, by contrast, identify abusive transactions on the basis of taxpayers' subjective purposes in undertaking them. The business purpose test and the step transaction doctrine are examples of such approaches. Finally, approaches that focus on policy tend to identify abuses by reference to underlying legislative intent. These approaches focus more on statutory interpretation and the legislative scheme than on factors such as taxpayers' purposes or the objective reasonableness of taxpayer decisions. None of the doctrine surveyed above focus on policy to the exclusion of form or purpose, but the U K trend toward submerging the step transaction doctrine into the ordinary process of statutory interpretation appears to be a move in this direction. 2 7 3 On Part IVA of the Australian Income Tax Assessment Act 1936, see Waincymer, supra note 10. New Zealand's general anti-avoidance rule is enacted as s. BG1 of the Income Tax Act, 2004. Sweden introduced, repealed and subsequently re-introduced a general rule. On the Swedish experience with the Tax Avoidance Act, SFS 1980:865, see Leif Muten, "The Swedish Experiment with a General Anti-Avoidance Rule" in Cooper, Tax Avoidance and the Rule of Law, supra note 10, 307. 96 Nothing suggests that the three common approaches to identifying tax avoidance are mutually exclusive. In fact, they are not. It is impossible to consider the economic reasonableness of taxpayers' choices of legal forms for transactions without considering whether there are substantial non-tax purposes behind those choices. Similarly, the reasonableness of the legal structures chosen to achieve economic outcomes provides an important window on taxpayers' purposes and evidence with which to evaluate the existence of non-tax purposes. Finally, courts considering tax avoidance transactions, even ones that are totally economically unreasonable and that serve no non-tax purpose, cannot avoid considering whether the tax results sought by taxpayers are consistent with legislative intent. Indeed, courts that fail to consider this question would be derelict in their most basic duty to interpret the legislation in accordance with Parliamentary intent. The third common feature of the various anti-avoidance doctrines is that, as Ward explains, they all reflect an attempt by the courts to "limit the rights of [individuals] in property, contract, corporate and other fields of law by balancing [individual] rights against the benefits to society that are achieved by such limitations." 2 7 4 Courts that have accepted the need for some type of general anti-avoidance doctrine acknowledge that there is a societal interest in balancing taxpayers' rights to plan their affairs with the consequences that unfettered tax planning would have for the fisc and for other taxpayers (again demonstrating the significant equity problems created by abusive tax avoidance). Courts in those jurisdictions have accepted that the courts, as a key social institution, have a legitimate role in forming and preserving this balance. They do not believe that their role is simply to administer and apply Parliamentary language without concern for the consequences of Ward, "Judicial and Legislative Approaches", supra note 113 at 115-16. 97 various interpretations in terms of Parliamentary intent and broader social values such as equity. Canada, of course, has enacted the G A A R in section 245 of the Act. Upon review of the various anti-avoidance doctrines applied in other jurisdictions, the reason for Parliament's decision to do so becomes clear: Canadian courts, following the lead of the Supreme Court of Canada, have consistently refused to apply any meaningful anti-avoidance doctrines not enacted in the legislation. The only extra-statutory doctrines applied in Canadian tax law are a narrow sham transaction doctrine, the ineffective transaction doctrine, and a legal substance doctrine that verges on the sham doctrine. In my view, none of these doctrines are anti-avoidance doctrines at all, since they would not apply to genuine, legally-effective transactions, which tax avoidance transactions invariably are. The remaining question is why Canadian courts have refused to accept the responsibility for combating avoidance that courts in other jurisdictions have assumed. The answer to this question lies in the Canadian approach to statutory interpretation in the tax context, and in the narrow view that Canadian courts have taken of their role in the tax system. Canadian courts have consistently indicated that it is for Parliament, not courts, to close the door to tax avoidance and to make the difficult decisions needed to preserve an equitable balance between individual tax planning interests and the societal interest in a fair distribution of the tax burden. Thus, any review of the G A A R must begin with the Canadian approach to statutory interpretation and the appropriate role of the courts in administering the tax system. 98 V. THE GAAR Given the absence of meaningful judicial anti-avoidance doctrines in Canadian tax law, any meaningful limits on tax avoidance must be found in the legislation itself. There are two kinds of such provisions in the Income Tax Act - a multitude of specific rules aimed at preventing tax avoidance in particular circumstances; and the G A A R , which provides overriding protection against abusive tax avoidance that otherwise escapes the myriad specific anti-avoidance rules. After many years of relying almost exclusively on detailed anti-avoidance amendments, Parliament concluded, for the reasons identified by the Royal Commission on Taxat ion 2 7 5 and acknowledged in Stubart, 2 7 6 that specific anti-avoidance rules alone are insufficient. Such rules are merely stopgap measures, introduced only after taxpayers have identified new ways to circumvent the existing rules. Furthermore, relying entirely on specific anti-avoidance rules leads to ever-increasing legislative complexity and obtuseness. This Chapter does not review the debate over whether a statutory general anti-977 avoidance rule is necessary, although I believe that one is. Rather, this Chapter analyzes the current G A A R and how it has been interpreted by the Supreme Court of Canada. The starting point for this analysis is a consideration of Canadian statutory interpretation generally, for two reasons. First, the G A A R itself was Parliament's response Report of the Royal Commission on Taxation, supra note 10 at 554-55. 276 Supra note 31 at 6324, where the Court spoke of the "action and reaction endlessly produced by complex, specific tax measures aimed at sophisticated business practices, and the inevitable, professionally guided and equally specialized taxpayer reaction." 2 7 Canada witnessed a vociferous debate over whether a GAAR, or even constitutional. For a review of the arguments for and against, see Arnold & Wilson, "Part II", supra note 96 at 1126-42. For an argument by the then Deputy Minister of Finance that a statutory GAAR was necessary, see Dodge, supra note 11. For the case against a statutory GAAR, refer to Howard J. Kellough, "A Review and Analysis of the Redrafted General Anti-Avoidance Rule" (1988) 36:1 Can. Tax J. 23. Such a debate is ongoing in the UK. For instance, refer to Judith Freedman, "Defining Taxpayer Responsibility: In Support of a General Anti-Avoidance Principle" [2004], no. 4 British Tax Review 332, and "Converging Tracks? Recent Developments in Canadian and UK Approaches to Tax Avoidance" (2005) 53:4 Can. Tax J. 1038. 99 to the manner in which Canadian courts were interpreting tax laws. Parliament intended the G A A R to mandate a break from the dominant interpretive tradition. Second, the courts' approach to the G A A R must inevitably be influenced by their overall approach to statutory interpretation. After reviewing the general approach to statutory interpretation in Canadian tax law, this Chapter wi l l focus explicitly on the G A A R . It wi l l review the structure of the G A A R , identifying and analyzing the main concepts used, such as tax benefit, the purpose test, series of transactions, and the misuse and abuse saving clause that has caused much uncertainty. Those concepts have not been invented in isolation, but they are related to and share much in common with the other anti-avoidance doctrines discussed in the previous Chapter. Those concepts also display elements of the three main approaches to identifying abusive tax avoidance discussed in Chapter 3, such as the purpose- and policy-based approaches. O f course, in addition to the provision itself, an analysis of the G A A R must 278 consider the Supreme Court of Canada's 2005 decisions in Canada Trustco and Matthew v. Canada,219 which dictate (for now) how Canadian courts must interpret the G A A R . A. Statutory Interpretation and Tax Avoidance in Canada The Supreme Court of Canada has consistently declined to endorse any extra-statutory anti-avoidance doctrine when interpreting tax legislation. Rather, the Court has repeatedly referred to taxpayers' right to arrange their affairs in order to reduce taxes; to the need to interpret tax legislation as strictly and literally as possible in the circumstances of any given case; and to the importance of respecting legal relationships, even in the case of Supra note 80. 2005 D.T.C. 5538. 100 artificial tax-motivated transactions. The reason why the Court has done so is founded, ultimately, in its approach to statutory interpretation. As Arnold explains, tax avoidance and statutory interpretation are intimately related 2 8 0 because tax avoidance involves questions of legislative purpose and intent, and the degree to which one believes that legislative purpose and intent are relevant varies according to one's approach to statutory interpretation. Therefore, since courts' approach to statutory interpretation significantly affects their approach to tax avoidance, it is appropriate to briefly review the Supreme Court of Canada's approach to statutory interpretation in the tax context. In Canadian tax law, statutory interpretation begins with Duke of Westminster. That decision laid down three key principles (the "Westminster principles") that have informed the Canadian approach to interpreting tax statutes ever since: first, that taxpayers have the right to arrange their affairs so as to reduce taxes; second, that tax legislation should be interpreted strictly or literally; and third, that the legal relationships resulting from transactions, not their economic substance, are determinative for tax purposes. Although the Supreme Court of Canada has ostensibly abandoned the second principle, that of literal interpretation of tax statutes, the Court has repeatedly rejected any serious challenge to the other two principles. Lord Tomlin formulated the first Westminster principle, acknowledging a right to tax planning, in the following terms: " . . .every man is entitled i f he can to order his affairs so as 282 that the tax attaching under the appropriate Acts is less than it otherwise would be." This Arnold, "Reflections", supra note 10 at 4. 1 Supra note 53. 2 Ibid, at 19. 101 principle has never been seriously challenged in Canadian tax law. In fact, even when the Supreme Court rejected the principle of strict interpretation as outdated in Stubart, the Court 984 took pains to reiterate the importance of this first principle. In its most recent discussion of interpreting tax statutes, the Supreme Court elevated this judicial principle to the status of 9RS a fundamental Parliamentary policy. The Court asserted that Parliament intends tax laws to be predictable, certain, and fair so that taxpayers can take advantage of tax benefits conferred by the Act. Therefore, taxpayer must be able to predict the tax consequences of their transactions as accurately and reliably as possible, which in turn requires a degree of certainty that generally results from preferring a literal approach to statutory interpretation. The third Westminster principle, that the legal forms of transactions are determinative for tax purposes, has occasionally been questioned, as in Bronfman Trust,286 but the Court has ensured its survival in the face of those challenges. Indeed, Shell is the Court's definitive statement that taxpayers' legal relations are determinative. 2 8 7 Only the second Westminster principle, that of strict interpretation, has been modified in Canadian law. Originally, strict interpretation meant that tax would be chargeable only if 988 a taxpayer came within the literal meaning of the tax statute. A n approach focusing on the This principle, or a variant thereof, is recognized in the tax system of most modern Western jurisdictions. However, the extent to which taxpayers are restrained from unfettered tax planning varies from jurisdiction to jurisdiction, as discussed in Chapter 4. 284 Stubart, supra note 31 at 6325, per Wilson J.: "I think that Lord Tomlin's principle is far too deeply entrenched in our tax law for the courts to reject it in the absence of clear statutory authority." 285 Canada Trustco, supra note 80 at 5529, para. 31. The Court made these comments in the context of constructing a competing tax policy interest to be balanced against attempts to prevent abusive tax avoidance. 286 Supra note 135. Refer to the discussion in Chapter 4 under the heading "Economic Substance - Substance and Form in the UK and Canada." 287 Supra note 99. As I discuss in Chapter 6, Canada Trustco presented the Court with an opportunity to give more weight to economic substance when interpreting the GAAR than it receives in applying the ordinary charging provisions of the Act. The Court declined to do so, emphasizing again the primacy of legal relationships and in the process significantly constraining the utility of the GAAR to combat abusive tax avoidance. 288 Duke of Westminster, supra note 53 at 24, endorsing Partington v. Attorney-General (1869) L.R. 4 (H.L.). Such a principle operated as a double-edged sword, however. Just as the Crown could not levy tax unless 102 plain meaning of the provisions was superficially appealing to courts seeking to interpret and apply tax laws to complex transactions, but proved to be no more than an open invitation to aggressive tax planners to engage in transactions that literally complied with the wording, but not the spirit, of the Act. In Stubart, the Court recognized that abusive tax avoidance is a problem with adverse consequences for the tax system in general, and even referred to that 289 problem as a reason for modernizing the Canadian approach to interpreting tax statutes. The Court therefore adopted the modern rule of statutory interpretation, which, when determining how to apply a statutory provision, considers the context of the provision relied upon by a taxpayer, the scheme of the Act as a whole, and the intent of Parliament in addition to the plain and ordinary meaning of the statutory language. 2 9 0 The Court explicitly chose this approach over a business purpose test as the best means of combating tax avoidance, speculating that this approach would capture "the conduct of [taxpayers] which clearly falls 291 within the 'object and spirit' of the taxing provisions." In addition to adopting the "object and spirit" approach, the Court laid down a series of interpretive guidelines intended to assist courts in dealing with tax avoidance. Those guidelines included a somewhat vague suggestion that the formal validity of transactions would not be sufficient to entitle taxpayers to tax benefits where the "object and spirit" of allowance or benefit provisions was defeated by "procedures blatantly adopted by [taxpayers] to synthesize a loss, delay or other tax saving device," even if those actions are taxpayers were clearly caught by the literal wording of the statute, taxpayers could not escape taxation on the basis that the statute was not intended to apply to them if, on a literal reading of the statute, they were taxable. 289 Stubart, supra note 31 at 6321. Also Hogg, supra note 107 at 564. 290 Stubart, ibid, at 6323. 291 Ibid, at 6322. not "artificial." The Court suggested that its interpretive guidelines were appropriate to reduce the cycle of tax avoidance and legislative amendment. Stubart was widely regarded as fundamentally changing the interpretation of tax law in Canada. It appears that by adopting the "object and spirit" approach, the Court intended to create a test that would identify characterize as abusive transactions that resulted in tax benefits not contemplated by the legislation. Such a purposive approach would supposedly have freed the courts from the confines of a literal approach under which every transaction received effect for tax purposes regardless of its purpose or whether consequent tax benefits were contemplated by Parliament. However, at the same time as ostensibly revolutionizing its approach to statutory interpretation, the Court reinforced the other two Westminster principles, namely the right to order one's affairs to reduce taxes and the primacy of legally valid relationships (subject only to exceptions for shams and for legally ineffective transactions). Furthermore, the Court's language suggested that, in the absence of statutory prohibitions, tax planning should be the norm: .. .where the substance of the Act, when the clause in question is contextually construed, is clear and unambiguous and there is no prohibition in the Act which embraces the taxpayer, the taxpayer shall be free to avail himself of the 291 beneficial provision in question. The reference to "clear and unambiguous" clauses and the need for a "prohibition in the Act which embraces the taxpayer" foreshadowed the Court's retreat from purposive to literal interpretation that occurred in a series of tax avoidance cases after Stubart,294 culminating in Shell. In fact, the Court never again mentioned the Stubart interpretive guidelines in any tax 292 Ibid, at 6324. At that time, section 137 of the Act prohibited deductions from income that, if allowed, would "unduly or artificially" reduce income. Section 137 was later renumbered to section 245, and was eventually replaced by the GAAR. 293 Ibid, at 6324. 2 9 4 Including, most importantly, Antosko v. R., 94 D.T.C. 6314 (S.C.C.), and Friesen v. R., 95 D.T.C. 5551 (S.C.C.). 104 decisions, notwithstanding the fact that the Court had apparently created those guidelines as a 295 response to the problem of tax avoidance. B y the time it decided Shell, the Court had returned to a literal approach to interpreting tax statutes, focused primarily on the "plain meaning" of their provisions. Under a plain meaning approach, clear and unambiguous statutory provisions must simply be given their literal effect, and considerations of their object and spirit are unnecessary. Object and spirit would be considered only if a provision were ambiguous. In Shell, McLach l in J. explained this approach in the following terms: .. .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. Where the provision at issue is clear and unambiguous, its terms must simply be applied. 9 6 The comments in Shell oversimplify the importance of plain meaning. In other recent decisions, the Court has indicated that interpretation should not focus on the plain meaning of 297 the language to the exclusion of all other factors such as context and legislative purpose. Nonetheless, the Court continues to emphasize the need to give significant weight to the literal words chosen by Parliament. Although the Court has not been consistent in the years since Stubart, its jurisprudence generally came to revolve around a text-based approach to the Act, limiting considerations of context and purpose to instances of ambiguity in the statutory provisions. W 5 Taylor, supra note 121 at 17:13. 296 Shell, supra note 99 at 5676, para. 40. 2 9 7 Such as in Singleton, supra note 89 at 5542, para. 68, in which LeBel J. ovserved that "if the 'plain meaning' approach is to make any sense at all, it surely cannot mean that we are always to ignore context when interpreting statutory language. Rather, it must be understood to say that although context is always important, sweeping considerations of general statutory purpose cannot outweigh the specific statutory language chosen by Parliament." In addition, the Court has taken a more overtly purposive approach in some cases, such as Will-Kare Paving & Contracting Ltd. v. /?., 2000 D.T.C. 6467, and 65302 B.C. Ltd. v. R., 99 D.T.C. 5799. 105 In its 2005 Canada Trustco decision, the Court again observed that tax statutes are no different than other statutes, and that tax provisions should "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." 2 9 8 The Court emphasized that all statutory provisions must be read "according to a textual, contextual and purposive analysis to find a meaning that is harmonious with the Act as a whole ." 2 9 9 The Court appears to recognize that the context of a provision - the overall statutory scheme in which the provision is found - and the purpose of a provision - identifying what Parliament intended to achieve by including it in the statute - are relevant in all cases. Such comments are reminiscent of Stubart, which itself should have resulted in less tolerance for tax avoidance. However, despite these encouraging comments, the Court once again emphasized that, in the context of a detailed tax statute, precise and unequivocal words wi l l play a dominant role in the interpretive process. The Court thereby again invites a predominantly textual approach to statutory interpretation, in which considerations of context and purpose are secondary and arise only when statutory language is ambiguous. 3 0 0 To justify a text-first approach, the Court observed that Parliament intends taxpayers to rely on clear and precise statutory wording in order to plan their transactions. The need for certainty therefore dictates that Canada Trustco, supra note 80 at 5526, para. 10. 299 Ibid, at 5526, para. 10. The Court appears to simply be rephrasing the modern rule of statutory interpretation rather than creating new interpretive guidelines. 300 Ibid, at 5526, para. 10. Such an approach recalls the language of earlier cases like Antosko and Friesen, both supra note 294, which mandated a plain meaning approach unless the provisions were ambiguous. It is too early to predict whether Canada Trustco heralds a new approach to interpreting tax statutes that places more weight on context and purpose in every case, nor merely in instance of textual ambiguity. Alternatively, Canada Trustco may simply provide a justification for reinforcing a focus on plain meaning and textual interpretation. 106 clear language in detailed provisions should not be qualified by "unexpressed exceptions derived from a court's view of the object and purpose of the provision." 3 0 1 From a tax policy perspective, a literal approach to statutory interpretation is highly problematic because it invites tax avoidance. It has proven remarkably easy for taxpayers to legally structure their affairs in ways that literally comply with the various requirements of the Act in order to obtain tax benefits that Parliament did not intend them to received when it enacted the legislation. The anti-avoidance doctrines reviewed in the previous Chapter illustrate circumstances in which courts have decided to depart from a literal interpretation in order to give effect to what the court believes to be the legislative intent. Canadian courts, however, have not been prepared to do so. Canadian courts have consistently espoused the view that if, on a literal reading, tax legislation does not apply as Parliament intended it to, then Parliament must fix the problem. According to this view, it is not for the courts to evaluate the tax policy consequences of competing interpretations and to make normative choices between those consequences. Such an approach is clearly displayed in Antosko. In Antosko, the taxpayers purchased shares of an indebted company from its creditor, a publicly-owned finance board. The board subsequently sold the taxpayers the debt owed to it, plus accrued but unpaid interest, for $10. The taxpayers then deducted the value of interest accrued on the debt prior to their purchasing i t . 3 0 4 However, as a publicly-owned entity, the finance board was exempt from tax on interest. A s a result, neither the board nor 301 Canada Trustco, supra note 80 at 5526, para. 12, citing 65302 B.C. Ltd., supra note 297, itself citing Hogg, supra note 107. 3 0 2 Taylor, supra note 121 at 17:7. 303 Supra note 294. 3 0 4 Subsection 20(14) of the Act allocates the portion of the accrued interest that relates to the period before a debt is transferred to the income of the transferor and permits the purchaser of the debt to deduct that amount from its income. 107 the taxpayers paid tax on the prior interest that had accrued while the board owned the debt, even though the taxpayers would ultimately receive that interest if the debt were repaid. The Crown sought to deny the taxpayers a deduction under subsection 20(14) on the basis that it resulted in a tax-free windfall that was not intended by Parliament. A l l parties agreed that subsection 20(14) exists to prevent double taxation by ensuring that previously accrued interest is allocated to the income of the transferor for the period during which that party held the debt. The Crown argued that subsection 20(14) not only existed to prevent double taxation but also to ensure that accrued interest was included in some party's taxable income, since the transferor may not have previously included in its income interest that it had not yet received. Because the finance board selling the debt was tax-exempt, it would not be taxed on the interest allocated to it under subsection 20(140). The Crown argued, therefore, that allowing the taxpayers a deduction for the prior accrued interest would create a tax-free windfall and would defeat Parliament's underlying intention of taxing interest income at least once. 3 0 6 The Court agreed that subsection 20(14) exists to prevent double taxation of prior accrued interest, but rejected the argument that subsection 20(14) also preserves single taxation. The Court observed that, since the finance board was always tax-exempt, none of the accrued interest would ever have been taxed. The Court concluded that Parliament must have anticipated such a situation when it created a tax-exempt status for entities like the 3 0 5 This view is premised on the transferor, as the previous owner of the debt, being legally entitled to receive the interest that had accrued up to the time the debt is transferred, which would presumably increase the consideration paid to them for the debt. They would receive value for the interest that they were entitled to receive in addition to the principal of the debt, which value the purchaser would presumably pay, and therefore it is appropriate to allocate the value of that prior accrued interest to the transferor and to allow the purchaser a corresponding deduction. 3 0 6 Furthermore, the debtor company owed the public board $5 million plus interest. The taxpayers acquired that debt, plus accrued interest, for just $10. Thus, it does not appear that the value of interest accrued while the board held the debt was reflected in the consideration paid for the debt, and the underlying rationale for preventing double taxation does not exist. 108 finance board. In the Court's view, subsection 20(14) was clear and unambiguous and dealt only with the allocation of accrued interest to the parties' taxable income. The Court did not believe that the consequences of such an allocation, in terms of whether tax would ever be TAT levied on that interest, were relevant to the application of the provision. The Court then stated that it was not for the courts to determine whether a transaction is "one which renders [a] taxpayer deserving of a deduction." 3 0 8 When the terms of a section are met, the Court held, a taxpayer may rely on it and it is up to Parliament to specifically prevent further reliance by amending the legislation. The Court stated that a normative assessment of the consequences of applying a provision in different ways is relevant only where the provision itself is ambiguous, not where its meaning is clear and plain. A n assessment of the desirability of the consequences of simply applying a clear and plain provision is for the legislature, not the courts. Antosko illustrates two defining features of the Canadian approach to statutory interpretation. The first is a continuing adherence to the tradition of strict interpretation of tax legislation. Stubart had indicated that it is appropriate for the courts to consider context and purpose when interpreting tax legislation. Considering context and purpose means that courts wi l l need to acknowledge the consequences of a literal interpretation and to decide whether, based on their understanding of Parliament's intent, the consequences of a literal interpretation are what Parliament intended in the circumstances of a particular case. In Antosko, the Court could have considered whether Parliament would have intended the prior 3 0 7 The Court also suggested that practical difficulties would be created for bond markets if taxpayers acquiring debts we re required to investigate whether the vendors were tax-exempt entities in order to know whether prior accrued interest would be taxable, and that this would affect the valuation of securities. With respect, this concern is specious, since it is common for vendors and purchasers of securities to consider and to investigate the tax status of securities and other parties, and to factor tax consequences into the valuation of securities. The practical problems that the Court feared are in fact already commonplace in a sophisticated marketplace. 308 Supra note 294 at 6320. 109 accrue interest to go untaxed. Instead, the Court marginalized the Stubart approach by confining purposive analysis to circumstances of ambiguity. Unfortunately, legislative ambiguity is often to be found in the eye of the beholder. Although, on a superficial reading, a legislative provision may appear clear and plain, ambiguity may become apparent when it is read in light of the overall legislative scheme and purpose. For instance, subsection 20(14) appears on its face to deal simply and plainly with the allocation of prior accrued interest to different taxpayers' incomes. However, when the overall scheme of the Act is considered, it could be that subsection 20(14) is premised on an assumption that both the transferor and the acquiror of a debt would be taxable on the interest, thus creating a double taxation problem for the provision to resolve. 3 0 9 Thus, in a situation like Antosko where the transferor is taxable on interest included in their income, the assumption of a double taxation problem is not borne out because no tax wi l l be paid on the portion of the interest allocated to the transferor. The provision therefore becomes ambiguous because it is not clear that Parliament intended such a no-tax outcome, as opposed to merely a "taxed once" outcome, to result. However, an approach focusing on the apparent "plain meaning" of the provision, and failing to consider context and purpose, would neither recognize such an ambiguity nor permit the Court to engage in a debate over which outcome is more desirable - the failure to tax an amount of interest at all or the denial of a deduction to the acquiror of the debt. After Antosko, the Supreme Court of Canada repeatedly referred to the plain meaning of tax legislation and was increasingly reluctant to engage in the kind of detailed, purposive 3 0 9 Furthermore, logic suggests that concerns about double taxation would arise only where either party would be taxable, not in circumstances where one party would not be taxable. Despite the Court's suggestion that Parliament must have contemplated debt transfers by tax-exempt entitles, Parliament may also have contemplated only double taxation circumstances when it enacted subsection 20(14). 110 310 interpretation that Stubart suggested should be the norm in tax cases. In fact, the Court never again looked to the purpose of a provision in order to assist with interpreting it in a tax-avoidance context. 3" In its most recent comments on statutory interpretation, the Court again emphasized the importance of a textual interpretation where the words are "precise and unequivocal," observing that in those instances the ordinary meaning of the words must play a dominant role in interpreting the statute.3 1 2 The second defining feature of Canadian statutory interpretation illustrated in Antosko is the courts' insistence that countering tax avoidance through purposive interpretation is not their proper role. The courts have repeatedly emphasized that it is up to Parliament alone to 313 deal with abusive tax avoidance, presumably through legislative amendments. The statement in Antosko that it is not for the courts to make normative judgments about the potential consequences of a literal reading of the Act illustrates Canadian courts' restrained view of their role in the tax system. In a series of judgments that followed Antosko -particularly Neuman,3H Duha Printers,315 Continental Bank,316 and Shell311 - the Court carved out a narrow, confined judicial role in the Canadian tax system. According to the 3 1 0 The Court's jurisprudence did not develop a plain meaning focus in a straightforward manner, but in fits and starts. In addition to cases endorsing a plain meaning approach, there were cases in which the Court referred to the scheme of the Act, the object and spirit of the relevant provisions, and legislative purposes. For instance, in Notre-Dame de Bonsecours v. Communaute Urbaine de Quebec et al, 95 D.T.C. 5017, decided just a year after Antosko, the Court endorsed a "ideological approach," which identifies the purpose of a legislative provision in light of its context and the legislative intent, and gives a provision a strict or liberal interpretation in light of its identified purpose. By contrast, in R. v. Province of Alberta Treasury Branches and the Toronto Dominion Bank, 96 D.T.C. 6245, decided a year after Notre-Dame de Bonsecours, the Court held that, unless there is doubt as to the meaning of the legislation or ambiguity as to its application, then statutory provisions must be applied in accordance with their plain meaning regardless of their object or purpose. In an interesting passage revealing the confusing inconsistencies in the Court's own thinking, Cory J. observed that, even when ambiguity is not apparent, "it is always appropriate to consider the scheme of the Act, the object of the Act, and the intention of Parliament" when determining the plain meaning of the statute. 3 1 1 Taylor, supra note 121 at 17:26. 3 1 2 Canada Trustco, supra note 80 at 5526, para. 10. 3 1 3 Taylor, supra note 121 at 17:35. 314 Neuman v. The Queen, 98 D.T.C. 6297 (S.C.C.). 315 Duha Printers (Western) Ltd. v. The Queen, 98 D.T.C. 6334 (S.C.C.). 316 Supra note 143. 317 Supra note 99. I l l Court, the role of a court deciding a tax dispute is merely to identify taxpayers' legal 318 relationships and to apply the legislation as it is written to the transactions in issue. The courts' responsibilities do not extend to normative judgments that characterizing a transaction or interpreting a provision in a particular way would be inconsistent with Parliament's underlying purposes. 3 1 9 Such considerations, which inform numerous anti-avoidance doctrines such as the business purpose test, the step transaction doctrine, and the economic substance analysis, have been banished from statutory interpretation in Canada. In the previous Chapter, I argued that the various anti-avoidance doctrines adopted in other jurisdictions shared a common recognition that there is a societal interest in constraining the scope of tax planning because of its detrimental effects on the public purse and on equity as between taxpayers. In Canada, the Supreme Court of Canada has explicitly rejected the suggestion that equity as between taxpayers is a relevant consideration: .. .absent a specific provision to the contrary, it is not the courts' role to prevent taxpayers from relying on the sophisticated structure of their transactions.. .on the basis that it would be inequitable to those taxpayers who have not chosen to structure their transactions that way. Thus, it is not surprising that no effective judicial anti-avoidance doctrines have emerged. Parliament's decision to enact the G A A R was a response to the Supreme Court of Canada's unwillingness in Stubart to endorse a general judicial anti-avoidance doctrine beyond the sham and ineffective transactions doctrines. Parliament believed that the only way to effectively control tax avoidance was to legislate a new rule that required the courts to undertake the kind of normative, policy-oriented analysis that the courts refused to build in to 3 1 8 Taylor, supra note 121 at 17:7. 319 Ibid, at 17:7. 320 Shell, supra note 99 at 5677, para. 45. This sentiment is a radical reversal from, and complete rejection of, the suggestion in Bronfman Trust that it would be inequitable, and thus undesirable, for tax liability to depend upon a taxpayer's "sophistication at manipulating a sequence of events to achieve a patina of compliance" with the legislation. See Bronfman Trust, supra note 135, at 5067. 112 the ordinary process of statutory interpretation. Subsequent decisions have vindicated Parliament's choice, as the Court has comprehensively reinforced the Duke of Westminster approach and has rejected attempts to import normative considerations into the interpretive process. Therefore, in Canada at least, tax avoidance must be controlled by the G A A R . B. Outline of the GAAR - Tax Benefit, Avoidance Transaction and Misuse or Abuse The G A A R creates a three stage test for identifying abusive tax avoidance. First, only transactions that result in a "tax benefit" could be subject to the G A A R . Generally, a tax benefit means a reduction, avoidance or deferral of tax payable under the Act. Second, the G A A R applies to transactions resulting in a tax benefit only i f those transactions qualify as "avoidance transactions." Avoidance transactions are defined as transactions that would result in a tax benefit unless they are undertaken or arranged primarily for another purpose. 3 2 ' The G A A R also includes a statutory step transaction test that requires each step in a series of related transactions to be undertaken primarily for a non-tax purpose. 3 2 3 Third, the G A A R only applies to avoidance transactions i f it is reasonable to consider that they result in a 194 misuse of the provisions of the Act or in an abuse of the Act as a whole. However, the legislation offers no guidance as to how to determine whether a transaction results in a misuse or abuse. Thus, the G A A R contains three tests to be met before it applies: a tax benefit, a purpose test for avoidance transactions, and a misuse or abuse test. The G A A R 321 Income Tax Act, supra note 7, ss. 245(1). 322 Ibid. para. 245(3)(a). 323 Ibid. para. 245(3)(b). 324 Ibid. ss. 245(4). In addition to provisions of the Act itself, the GAAR applies to transactions that result in a misuse or abuse of the Income Tax Regulations, the Income Tax Application Rules, a tax treaty, or any other enactment that is relevant in computing tax under the Income Tax Act; or in an abuse having regard to those provisions read as a whole. Initially, the GAAR referred only to provisions of the Act. 113 represents Parliament's attempt to define abusive tax avoidance while not interfering with legitimate tax mitigation, and to balance protection of the tax base with taxpayers' legitimate expectations to plan their affairs with a reasonable degree of certainty. Where transactions qualify as abusive tax avoidance subject to the G A A R , their tax consequences must be redetermined, as is reasonable in the circumstances of each case, in order to deny the resulting tax benefit. 3 2 5 The elements of the G A A R - tax benefit, avoidance transaction, and a saving clause using the misuse or abuse test - raise a number of questions in the course of interpreting and applying the G A A R , such as how to identify tax benefits, what purposes should be looked at in identifying avoidance transactions, and what it means to say that transactions result in a misuse of a statutory or regulatory provision or an abuse of those provisions read as a whole. C. Identifying a Tax Benefit Only transactions that result in a tax benefit might qualify as avoidance transactions. The starting point in every G A A R analysis, therefore, is to determine if the transaction under review results in a tax benefit. Subsection 245(1) defines tax benefit, in part, as a "reduction, avoidance or deferral of tax or other amount payable under [the] Act or an increase in a refund of tax or other amount under [the] Ac?." Thus, the definition of "tax benefit" is significantly broader than the phrase "tax avoidance" might suggest. In its simplest sense, 325 Ibid. ss. 245(2). In addition, subsection 245(5) lists a number of adjustments that may be made to redetermine the tax consequences to deny the tax benefit under subsection 245(2). 3 2 6 The remainder of the definition of "tax benefit" refers to transactions that result in reductions of amounts that would otherwise be payable under a tax treaty, and was enacted to deal with concerns that the GAAR would not be applicable to transactions that qualified taxpayers for tax exemptions under tax treaties as opposed to under the Act itself. In this thesis, I will not consider whether the GAAR applies to treaty benefits. That issue deserves its own discrete analysis, as it raises important questions of international law regarding the interaction between domestic laws and treaties. 114 "tax avoidance" means undertaking transactions that result in no tax being payable when tax would have been payable had the transaction not been undertaken (i.e., had the taxpayer either done nothing or undertaken a more "natural" alternative transaction). For instance, in Imperial Oil,327 Imperial O i l avoided tax under Part 1.3 of the Act by making short-term loans over its year-end to bank subsidiaries using funds that it typically invested in assets subject to tax. 3 2 8 If, on the other hand, Imperial O i l had either done nothing and held the funds itself at its year-end, or had continued to invest the funds in taxable assets as it did throughout the remainder of the year, it would have been taxed on those funds. In addition to transactions resulting in no tax at all, however, the G A A R also applies to transactions that result in other kinds of tax advantages, such as reductions and deferrals of tax. 3 2 9 Many of the cases involving the G A A R have concerned reductions and deferrals of tax, as opposed to outright avoidance. For instance, in many cases taxpayers have structured transactions to receive income in lower-taxed forms, such as by extracting corporate surplus in the form of capital gains instead of dividends. Such transactions, known as "surplus strips," are attractive because the capital gains income received is taxed at a lower rate than dividends. 3 3 0 In addition, in other cases taxpayers have acquired assets at little or no 331 economic cost for the purpose of claiming capital cost allowance. Capital cost allowance typically results in a tax deferral because the rates at which taxpayers are entitled to claim the iZI Supra note 108. 3 2 8 The facts of this case were discussed in Chapter 3 under the heading "Common Approaches to Identifying Tax Avoidance - Policy." 3 2 9 Arnold & Wilson, "Part II", supra note 96 at 1154. 3 3 0 In addition, under section 110.6 of the Act, taxpayers disposing of shares of "qualified small business corporations" are entitled to a tax exemption for the first $500,000 of capital gains realized in their lifetime. As a result, many surplus strips also involve a complete or partial avoidance of tax in addition to a reduction, since converting what would otherwise be dividend income from such corporations results in the receipt of tax-exempt capital gains. 3 3 1 This result is achieved when a taxpayer uses borrowed money that is cycled back to the lender so that it is effectively repaid, or uses non-recourse financing that does not expose the borrower to any economic risk. 115 allowance are usually higher than the actual rate of depreciation of their business assets, resulting in lower taxes in early years and recapture that increases income in future years when the assets are disposed of. Tax on the value of the capital cost allowance is thereby deferred until the year in which those amounts are recaptured into income. 3 3 2 Thus, the G A A R is a broad rule that could potentially apply to a wide variety of tax-avoiding, tax-reducing, and tax-deferring transactions. Given the wide variety of taxpayers' transactions, and the variety of ways in which taxpayer may reduce their taxes, it seems appropriate to define "tax benefit" so broadly. Notably, the Act is silent about the quantum of tax benefits. Therefore, the G A A R applies equally to transactions resulting in tax reductions of $1 and $1 mill ion. The main issue that arises in connection with identifying tax benefits is that the phrase "reduction, avoidance or deferral of tax" necessarily means a reduction, avoidance or deferral in relation to what would have happened had a transaction not been undertaken. Identifying a tax benefit therefore requires a notional comparison with imaginary "proper," or normative, tax consequences that would have resulted under different, hypothetical, circumstances. 3 3 3 There is no real question that identifying a tax benefit requires some kind of comparison; however, there is some uncertainty as to which alternative scenario should be used as a comparator. M a y one assume that the normative tax consequences are those that would have resulted had the tax-reducing transaction not occurred and the taxpayer done nothing else in its place, or should one assume that the taxpayer would have entered into a different transaction rather than doing nothing at all? The courts have concluded that the 3 3 2 Under section 13 of the Act, there is no recapture of previously-claimed capital cost allowance until all assets of a particular class are disposed of. Therefore, by keeping at least some property in a particular class forever, a taxpayer could permanently defer tax by never triggering recapture. 3 3 3 Arnold & Wilson, "Part II", supra note 96 at 1154. Also Duff, supra note 121 at 172. 116 114 answer to this question is that it depends on the circumstances. In my view, this position is reasonable and entirely appropriate. It is also consistent with good tax policy. In some cases, it is appropriate to assume that taxpayers would have entered into different transactions, and to use those transactions for purposes of determining whether the transactions under review resulted in tax benefits. In such cases, the question would effectively be whether the taxpayers have engaged in legitimate tax mitigation. 3 3 5 In other cases, it is more appropriate to assume that taxpayers would have done nothing at all, and to use that scenario for comparison purposes. For instance, where it is reasonable that taxpayers would have entered into different transactions than the tax-reducing ones, it is appropriate to identify the tax benefits by comparison with the tax consequences of those alternative transactions. Such would be the case where taxpayers seeking desired economic results structured their transactions in particular way primarily to achieve tax reductions. It would be unreasonable to assume that taxpayers pursuing real economic outcomes, such as return on an investment or obtaining borrowed capital for use in their business, would have elected not to do anything else had they not entered into the tax-reducing transactions. To continue the Imperial Oil example, Imperial O i l regularly invested surplus cash in order to make a return, and followed a pattern of investing its funds primarily in assets that did not qualify for the exemption from Part 1.3 tax. In those circumstances, it is reasonable to conclude that Imperial O i l would otherwise have continued to invest its funds in non-exempt assets in the same proportion as its historical pattern, and to use that alternative hypothetical scenario as the comparator in order 334 Canada Trustee supra note 80 at 5527, para. 20 3 3 5 As discussed in Chapter 2, legitimate tax mitigation is either a transaction that results in the economic outcome that Parliament would have intended a taxpayer to realize in order to qualify for a tax reduction, or a transaction that takes advantage of a tax incentive that Parliament enacted in order to encourage taxpayers to engage in particular behaviour even without non-tax purposes. 117 to determine whether the loans in question resulted in a tax benefit. In Imperial Oil, a tax benefit did result because the short-term loans to bank subsidiaries resulted in no tax payable when, had the loans not taken place, it is reasonable to conclude that Imperial O i l would have invested the majority of the funds in assets that did not qualify for the exemption from Part 1.3 tax. 3 3 6 337 338 Similarly, a weak currency loan situation, such as Shell and Canadian Pacific, involves taxpayers who need borrowed capital for business purpose and structure their borrowing through an intermediate currency for tax purposes. Since the taxpayers' overriding motivation is to obtain borrowed funds, it would be unreasonable to assume that, absent the intermediate currency borrowing, the taxpayers would not have borrowed at all. Therefore, the more appropriate comparison is with direct borrowing in the desired currency. Comparison of the two borrowing structures makes it clear that weak currency borrowing results in a tax benefit. On the other hand, where tax-reduction transactions appear to be isolated transactions entered into primarily to obtain their beneficial tax results, and not merely a choice between more and less tax-efficient methods of achieving desired economic results, it is reasonable to 3 3 6 In its reasons for judgment in Imperial Oil, the Tax Court concluded "with no hesitation" that there was a tax benefit because the loans reduced the tax payable under Part 1.3 of the Act. Imperial Oil, supra note 108 (T.C.C.) at 1965. The Tax Court did not even consider whether this question required a comparative analysis and, if so, what the appropriate comparator was. It appears that this issue may not have been raised. Therefore, the Court appeared comfortable treating the loans more like the transaction in Canada Trustco, supra note 80, which was more of an isolated transaction where it could not reasonably be said that the taxpayer would have entered into an alternative transaction as opposed to doing nothing at all had the tax-reducing transaction not proceeded. Although I agree with the Court's conclusion that there was a tax benefit, I believe that, from a tax policy perspective, it is preferable to acknowledge that had Imperial Oil not made the loans in issue, it reasonably would have invested the same funds in other investments rather than doing nothing. This approach has the advantage of acknowledging that taxpayers pursuing valid commercial objectives will sometimes choose the least-taxed method of achieving them without treating the ultimate commercial objective as justification for any tax-motivated steps, as the so-called "choice" principle referred to in Chapter 3 suggests. This comparative approach also focuses the GAAR analysis more explicitly on distinguishing between tax avoidance and legitimate tax mitigation. 337 Supra note 99. 338 Supra note 99. 118 assume that the taxpayers would not have entered into different transactions had the tax-reducing transactions not occurred. Rather, it is reasonable to assume that the taxpayers would have done nothing had they not entered into the transactions that they did, and therefore the appropriate comparison for purposes of identifying tax benefits is with the tax consequences that would have resulted had the transactions not occurred. Such was the case in Canada Trustee?39 and Matthew.340 In Canada Trustco, Canada Trustco sought out a sale-leaseback transaction in order to obtain capital cost allowance that could be used to offset other leasing income. 3 4 1 Canada Trustco therefore paid an arranger to locate and arrange a transaction in which, using approximately $25 mil l ion of its own and $97 mill ion of borrowed funds, Canada Trust purchased assets from a U S vendor for $120 mill ion and leased the assets back to an intermediary who then sub-leased the assets back to the original vendor. 3 4 2 Immediately thereafter, the vendor prepaid the sub-lease, totalling $116.4 million, back to the intermediary, which then deposited $97 mill ion with the bank that had originally financed Canada Trustco's purchase and the remaining $19 mil l ion in a guaranteed government bond that would mature at $33 million. Canada Trustco then instructed the bank to use the $97 mill ion on deposit to offset its own finance payments as they came due, and the intermediary pledged the bond to Canada Trustco as security for its obligations under the 33y Supra note 80. 340 Supra note 279. 3 4 1 A sale-leaseback involves purchasing assets from a vendor that is using them in its business but cannot deduct capital cost allowance because they are in a loss position. Taxpayers with profitable leasing activities will be interested in purchasing the vendor's assets in order to access the capital cost allowance deductions. In a sale-leaseback, the vendor will typically sell the assets and then lease them back if they can obtain a better finance rate through leasing than they were paying to own the assets outright. The purchaser is willing to offer a lower rate because they are also benefiting from deducting capital cost allowance. Sale-leasebacks are therefore a common form of financing deal. In addition, however, a sale-leaseback transaction permits the original owner of the assets to convert the value of the potential capital cost allowance that it could not otherwise deduct into a saleable commodity, and permits the purchaser to effectively purchase tax deductions. In this way, a sale-leaseback can also be viewed as a tax-motivated transaction in which the parties trade on their relative tax positions. 3 4 2 Roughly $2.5 million went to the arranger as a fee, accounting for the difference between the $122.5 million assembled by Canada Trustco and the $ 120 million paid to the vendor. 119 head lease. A t the end of the series of transactions, therefore, the bank held and was entitled to the $97 mill ion that it had originally loaned out, and Canada Trustco was pledged a government bond that would mature at $33 mill ion to secure its initial $25 mill ion outlay. A s the new owner of the assets, Canada Trustco deducted capital cost allowance of $36 and $46 mill ion in the year of acquisition and the following year, respectively. 3 4 3 The sale-leaseback transaction resulted in a tax benefit to Canada Trustco because it entitled Canada Trustco to claim capital cost allowance (a tax deferral, i f not an outright tax reduction) that it could not otherwise have c la imed. 3 4 4 The facts did not suggest that Canada Trustco would reasonably have entered into some alternative transaction had it not chosen the sale-leaseback that it actually undertook. Rather, the facts indicated that Canada Trustco set out to locate a tax-advantageous transaction and that there was no overriding commercial motive for it. Therefore, it should not be necessary to consider a hypothetical alternative transaction in order to determine whether the sale-leaseback resulted in a tax benefit. In actual fact, the courts did not believe that a hypothetical comparator transaction was necessary. Because the sale-leaseback transaction was so strongly tax-motivated, the courts concluded that it could simply be compared to the tax consequences that would have resulted had Canada Trustco not entered into any transaction at a l l . 3 4 5 The transaction therefore 3 4 3 All of the preceding facts are based on the findings of Miller J. of the Tax Court of Canada in Canada Trustco, supra note 80 (T.C.C.). 3 4 4 Both Canada Trustco and the Crown agreed that the capital cost allowance resulted in a tax deferral because Canada Trustco would have had recapture when the assets were ultimately sold. However, the tax benefit might more appropriately be characterized as a simple reduction rather than a deferral. Canada Trustco purchased trailers because trailers were exempted from the restrictions on specific leasing property found in the capital cost allowance regulations. The Tax Court explained that trailers are exempted from the specified leasing property restrictions because the capital cost allowance rate supposedly reasonably approximates the actual rate of depreciation of the trailers. If the rate does in fact reasonably approximate the actual rate of depreciation, then there would not have been any recapture when the trailers were sold. Recapture arises only if the capital cost allowance claimed over time exceeds the actual depreciation of an asset. If there is no recapture, then there would not have been any deferral of tax. Rather, there would have been a straightforward reduction of tax. 345 Canada Trustco, supra note 80 (T.C.C.) at 599. Interestingly, at the Tax Court, Canada Trustco challenged this proposition, arguing that a comparator transaction is always needed in order to identify any tax benefit, and 120 resulted in a tax benefit because it entitled Canada Trustco to a deferral (or to a reduction) of tax that would otherwise have been payable had the sale-leaseback not taken place. The Matthew case indicates even more strongly that a hypothetical alternative transaction need not always be identified in determining whether transactions result in a tax benefit. In Matthew, a liquidator liquidating the Standard Trust Company ("Standard Trust'), an insolvent company holding a portfolio of mortgages with $52 mil l ion of unrealized losses, implemented a tax plan that transferred the mortgages to a new non-arm's length partnership ("Partnership A " ) at their historical cost, obtaining a 99 percent interest in Partnership A . Standard Trust then sold its interest in Partnership A to an arm's length party that itself sold the interest in Partnership A to a new partnership ("Partnership B") that included Matthew and other outside investors who together paid only approximately $1.5 mill ion for their partnership units. 3 4 6 When the mortgage portfolio assets were sold and write-downs taken, Partnership A realized losses of $52 mill ion and allocated 99% of those losses to Partnership B . Partnership B then allocated the losses to its own investors, who deducted millions of dollars of partnership losses from their own incomes. The transactions effectively transferred $52 mill ion of unrealized losses from the insolvent original company to new, arm's length that the appropriate comparator was an ordinary sale-leaseback transaction that did not involve borrowed funds that were immediately repaid. The Tax Court did not accept that argument, for good reason. That argument assumes that Canada Trustco would have entered into some form of sale-leaseback in any event, and that it was the borrowed funds and immediate repayment that distinguished the actual transaction from what reasonably would have occurred otherwise. The appropriate comparison, however, is between a sale-leaseback and no transaction at all because the cycling of borrowed funds did not in any way result in a tax deferral; rather, it was the purchase of assets itself that resulted in the deferral. The cycling of borrowed funds is relevant only at the later stage of determining whether, under subsection 245(4) of the Act, the transaction resulted in a misuse or. abuse of the Act. 3 4 6 The intermediate purchaser was OSFC Holdings Ltd. ("OSFC"), which obtained a 24% interest in Partnership B and therefore still claimed 24% of the losses through its interest in Partnership B even though it no longer held the Partnership A interest directly. OSFC's claim for a deduction was the subject of a separate case, OSFC Holdings Ltd. v. R., 2001 D.T.C. 5471 (F.C.A.) [OSFC Holdings]. As in Matthew, the GAAR applied and OSFC was denied the loss deduction. 121 investors who paid only $1.5 mil l ion for units of Partnership B , but shared $52 mill ion in losses deductible from other income. 3 4 7 The Matthew transactions depended on subsection 18(13) of the Act, which prevents a taxpayer whose ordinary business includes lending money from deducting a loss on the disposition of a mortgage if, within 30 days before and 30 days after the disposition, the mortgage is owned by a non-arm's length partnership. In those circumstances, the loss is added to the acquiring partnership's cost of the mortgage. Absent subsection 18(13), therefore, Standard Trust would have realized the $52 mill ion in losses itself on the disposition of the mortgage portfolio resulting from the transfer to Partnership A . To an insolvent company like Standard Trust, those losses would have no value. To outside investors with taxable income to shelter, the losses did have value. Accordingly, Standard Trust transferred the losses to Partnership A , in which those investors could acquire units in order to access the losses. 3 4 8 The parties ensured that Partnership A was non-arm's length to Standard Trust so that subsection 18(13) would apply and add those unrealized losses to Partnership A ' s cost of the mortgages. Then, once Partnership A had realized the losses by disposing of or writing down the mortgages, its members, which included Matthew and the other investors in Partnership B , would be entitled to deduct their proportionate share of those losses due to the flow-through nature of partnership losses under the Act. Therefore, Matthew and the other investors were wil l ing to purchase units of Partnership B , and even to fund a $5 mill ion payment to Standard Trust, in order to access proportionate shares of the $52 mill ion in losses to be realized from Standard Trust's mortgage portfolio. 3 4 7 This summary of the facts in Matthew is based on the decision rendered by the Supreme Court of Canada, supra note 279 at 5538-41. 3 4 Presumably, interests in Partnership A would be marketed for sale so that some value would be received for the losses transferred from Standard Trust. 122 It is clear that the transactions resulted in a tax benefit for the investors in Partnership B . The series of transactions, beginning with the creation of Partnership A and transfer of Standard Trust's mortgages, continuing with the sale of 99% of Partnership A to Partnership B , and finishing with the sale of units of Partnership B to the taxpayers, resulted in the taxpayers deducting losses that they would not otherwise have had, thus creating a tax reduction. 3 4 9 It is not necessary to compare these transactions to some other transactions that the investors might have undertaken in order to identify a tax benefit. It is not reasonable to assume that the taxpayers would have entered into some other transaction had they not engaged in these transactions, since these transactions were obviously constructed to transfer unrealized losses out of a company that had no use for them and to market them to outside investors who would pay in order to access a share of those losses. Unlike Imperial Oil and the weak currency cases, this case raises no question of whether the taxpayers had engaged in legitimate tax mitigation in pursuit of some underlying economic objective. Rather, the entire series of transactions was created to achieve a tax result for the investors and no other economic objective. Therefore, it is appropriate to judge whether the transactions resulted in a tax benefit by reference to what the taxpayers' tax situation would have been if they had not entered into any transactions at all. L ike Canada Trustco, Matthew involved isolated transactions undertaken for tax purposes and not in pursuit of any economic outcome other than to capture partnership loss deductions. In fact, that the transactions resulted in a tax benefit was not even seriously challenged in Matthew. 123 D. Avoidance Transactions and "Series of Transactions" Not all transactions that result in a tax benefit should be considered abusive tax avoidance. Some may merely be legitimate tax mitigation. Sound tax policy therefore dictates that a general anti-avoidance rule should not apply to all transactions resulting in a tax benefit. The G A A R does not do so. Rather, the G A A R applies only to transactions 350 resulting in a tax benefit if those transactions qualify as "avoidance transactions." Once it is determined that a transaction results in a tax benefit under subsection 245(1), the next step in the G A A R analysis is therefore to determine whether the transaction is an avoidance transaction. Thus, the definition of "avoidance transaction" is one of the provisions through which Parliament delineates between legitimate tax mitigation and abusive tax avoidance. The G A A R creates this distinction by using a purpose test to identify abusive tax avoidance, coupled with a form of step transaction test applying the purpose test to series of transactions. The first branch of the definition, paragraph 245(3)(a), defines an "avoidance transaction" as: .. .any transaction that, but for [section 245] would result, directly or indirectly, in a tax benefit, unless the transaction may reasonably be considered to have been undertaken or arranged primarily for bona fide purposes other than to obtain the tax benefit. Through the G A A R , therefore, Parliament identifies as abusive tax avoidance only transactions that it is not reasonable to conclude were undertaken primarily for a bona fide purpose other than to reduce, defer or avoid tax. 3 5 0 Avoidance transactions that nonetheless do not reasonably result in a misuse or abuse are not subject to the GAAR. See the subsequent discussion under "An Exception for Misuse, Abuse and Frustration." 124 It is clear that, through its definition of "avoidance transaction," the G A A R adopts 351 primarily a purpose-based approach to identifying abusive tax avoidance. The purpose test in subsection 245(3), like all purpose-based approaches to identifying abusive tax avoidance, raises certain questions that must be addressed in order to arrive at a workable test for identifying tax avoidance. First, what exactly is meant by the term "purpose?" The answer to this question requires distinguishing between taxpayers' purposes in entering particular transactions and the overriding motives in pursuit of which they arrange those transactions, which might not be the same things. Such a distinction is particularly important when considering series of transactions, in which particular steps may not have a non-tax purpose but are nonetheless undertaken in the course of pursuing taxpayers' ultimate non-tax motives. The question of purpose is further complicated by the fact that the definition of avoidance transaction speaks of what may "reasonably be considered" to have been the purpose of the transaction. Such language suggests that transactions' purposes are an objective matter, which contrasts with the subjective nature of taxpayers' motives or intentions. Second, subsection 245(3) raises the question of which purposes are permissible and which are not. B y referring to "a bona fide purpose other than to obtain the tax benefit," Parliament has chosen different wording than a traditional business purpose test. Therefore, Parliament must have intended the avoidance transaction test to operate somewhat differently from a pure business purpose test like the one discussed in Chapter 4. The third question 3 5 1 Refer to the discussion in Chapter 3 of the three main approaches - form, purpose, and policy - to identifying abusive tax avoidance. As I will discuss below under "An Exception for Misuse, Abuse and Frustration," the exception in subsection 245(4) is a type of policy-oriented approach. In my view, the combination of a purpose test in the definition of "avoidance transaction" with a policy-based exception represents an attempt by Parliament to address the inherent weaknesses of each approach, as discussed in Chapter 3. 3 5 2 Some of these questions are identified in Chapter 3. In this paper, I will merely explore how the GAAR itself and Canadian GAAR jurisprudence address certain issues, as well as whether, from a tax policy perspective, the current Canadian approach is appropriate. This paper does not propose to exhaustively identify all such questions that arise in connection with a purpose-based approach to tax avoidance. 125 arising is what is meant by the reference to the "primary" purpose of a transaction. This question requires acknowledging that transactions may serve multiple purposes, and that the competing purposes must be weighed because not all transactions with an element of a tax purpose should be considered abusive. i . Objective Purposes as Compared to Subjective Motives The purpose test is the key to the G A A R framework for identifying abusive tax avoidance. Indeed, as Arnold & Wilson observe, a non-tax purpose is "the essence" of an avoidance transaction. 3 5 3 Therefore, in order to determine whether a transaction is an avoidance transaction, it is essential to properly identify its purpose. In circumstances where tax-reducing transactions are not undertaken in the course of pursuing an overriding business or economic objective other than tax-reduction, identifying the transaction's purpose may be a relatively straightforward endeavour. For instance, in Matthew, the taxpayers acquiring units of Partnership B did not intend to carry on business through the partnership; they intended merely to access losses that could be deducted from their other income. In addition, the liquidator seeking to realize value from Standard Trust's mortgage portfolio did not transfer those mortgages to Partnership A with the intention that the partnership would carry on a business of holding the mortgage portfolio; the liquidator merely intended units of the partnership to be purchased by taxpayers looking to access the unrealized losses added to the partnership's cost of the mortgage portfolio under subsection 18(13). Since there is no indication that those transactions were undertaken in the context of pursuing a non-tax objective, it is not particularly difficult to identify their purpose as being to obtain a tax 3 5 3 Arnold & Wilson, "Part II", supra note 96 at 1155. 126 benefit. Their purpose was to transfer unrealized losses from one taxpayer to others through the use of subsection 18(13) and the flow-through nature of the partnership taxation provisions, so that the losses could be deducted by taxpayers with positive income. However, the purpose inquiry becomes more complicated when taxpayers arrange specific transactions for tax purposes in the context of pursuing overriding non-tax goals or objectives. In such circumstances, it is important to distinguish between the purpose for which a transaction is entered into, and a taxpayer's motivation in doing so, because those are not necessarily be the same th ing. 3 5 4 For instance, in a weak currency loan situation, the intermediate borrowing in a third currency is undertaken in the course of pursuing an overriding economic objective or goal, the raising of capital. However, the purpose of the intermediate borrowing transaction itself is to obtain beneficial tax consequences resulting ICC from a greater interest deduction. In these circumstances, the G A A R analysis requires distinguishing between the purpose of the avoidance transaction itself and the taxpayer's motivation in entering into the transaction. Although the intermediate borrowing in a weak currency borrowing situation is arranged in pursuit of a taxpayer's overriding objective of raising capital, the purpose of the intermediate borrowing itself is primarily, i f not solely, tax-reduction. The distinction between purpose on one hand, and motivation or intent on the other hand, is sometimes difficult to draw, but such a distinction is essential to properly applying the G A A R . In part, this distinction can be made on the basis that the test for an avoidance 3 5 4 Brian Arnold, "The Canadian General Anti-Avoidance Rule" in Cooper, Tax Avoidance and the Rule of Law, supra note 10, 221 at 231-32 [Arnold, "Canadian General Anti-Avoidance Rule"]. Also see Harry Erlichman et al., "The Statutory Context of the GAAR" in Harry Erlichman, ed. Tax Avoidance in Canada: The General Anti-Avoidance Rule (Toronto: Irwin Law Inc., 2002) 5 at 32-33. 3 5 5 In Shell and Canadian Pacific, both supra note 99, the taxpayers also sought to guarantee a foreign exchange gain resulting from locking in forward exchange contracts in a declining interest rate environment. The foreign exchange gain could be used to offset the higher interest deduction and, in fact, would be a tax-efficient way of doing so because 25% of the offsetting capital gain would have been tax-free. 127 transaction hinges on the purpose for which the transaction was arranged, which focuses on the purpose of the transaction as opposed to the taxpayer's motivation in entering into it. Transactions have purposes, in the sense that they have a function - they are designed to bring about a particular result. 3 5 6 When the result that a transaction is designed to bring about is a tax reduction or deferral, then the "purpose" of the transaction may be said to be to obtain a tax benefit. 3 5 7 For instance, taxpayers who need U S dollars for use in business do not borrow Australian dollars because they need U S dollars, but because first borrowing in Australian dollars results in reduced taxes as compared to borrowing U S dollars directly. The intermediate borrowing thus serves a tax function and obtains a tax result. Therefore, the purpose of the intermediate borrowing is not to raise capital but to bring about a particular tax reduction, although the intermediate borrowing is nevertheless ultimately motivated by a need for business capital . 3 5 8 The above analysis rests on the view that it is possible, objectively, to identify the purpose of a transaction by identifying a result that the transaction is intended to bring about. This analysis also rests on the premise that the purpose of a transaction is not necessarily 3 5 6 Erlichman et al., supra note 354 at 33, refer to the Oxford English Dictionary, which indicates that "purpose" is commonly defined as "that which one sets before oneself as a thing to be done or attained, the object which one has in view;.. .the object for which anything is done or made, or for which it exists, the result or effect intended or sought, end, aim." Similarly, Arnold & Wilson, "Part II", supra note 96 at 1157, cite the Privy Council decision in Newton v. Commissioner of Taxation, [1958] A.C. 450 at 465, in which the Privy Council held that "the word 'purpose' means not motive but the effect which it is sought to achieve - the end in view..." in support of their contention that subsection 245(3) establishes an objective purpose test testing the purpose of the transaction, not the taxpayer's purpose. 3 5 7 Arnold & Wilson, ibid, at 1157, contend that "the purpose of a transaction must be determined by reference to what the taxpayer did and the legal, commercial and tax results of what he did, rather than his motives and intentions." 3 5 8 In Canadian Pacific, supra note 99, the Federal Court of Appeal concluded that the GAAR did not apply to weak currency borrowing on the basis that the initial borrowing of Australian dollars was not an avoidance transaction because the purpose of that borrowing was to acquire working capital for use in the company's business. The Court appears to have confused the taxpayer's objective or motivation with its purpose for undertaking the intermediate borrowing itself, which was, objectively, to create a greater tax deduction than would otherwise have resulted from directly borrowing US dollars. In doing so, the Court misapplied subsection 245(3) of the Act. 128 identical to the taxpayer's motivation in arranging the transaction, since the taxpayer may undertake a transaction that serves a tax purpose in the course of pursuing other non-tax objectives (of course, i f a taxpayer pursued only tax objectives, then there would be no doubt that the transaction is an avoidance transaction). There is debate, however, over whether it is appropriate to ascribe a different purpose to a transaction than the taxpayer's motivation in arranging the transaction. For instance, Couzin notes that the definition of avoidance transaction refers not to the purpose of the transaction, but "to the purpose for which the transaction was undertaken or arranged." 3 5 9 He argues that concepts such as motive, intention, and purpose are not easily distinguishable and that the purpose of a transaction is not the result that the transaction is objectively designed to achieve, but rather the purpose or goal for which the person arranging the transaction has done so . 3 6 0 Therefore, in his view it would be inappropriate to speak of a transaction's purpose in the objective sense that the transaction brings about a particular result, at least without also considering a taxpayer's ultimate goal. With respect, I do not agree that Couzin's approach is the appropriate interpretation of the purpose test in subsection 245(3) for two reasons. First, the provision couches the purpose test in objective language by referring to whether the transaction "may reasonably be considered" to have been undertaken primarily for purposes other than to obtain the tax benefit. 3 6 1 B y using the language "may reasonably be considered," Parliament clearly adopted an objective test that asks "what a reasonable taxpayer would have considered the primary purpose of the transaction to be, not what the particular taxpayer's motive or 3 5 9 Robert Couzin, "Subsection 245(3): A Framework," in Report of the Proceedings of the Forty-Ninth Tax Conference, 1997 Conference Report (Toronto: Canadian Tax Foundation, 1998) 4:1 -15 at 4:9. 360 Ibid, at 4:10-11. 3 6 1 Arnold & Wilson, "Part II", supra note 96 at 1157. 129 subjective intention.. .was." 3 6 2 In Canada Trustco, the Supreme Court of Canada confirmed that Parliament intended this test to be objective, asking what the purpose of the transaction reasonably was . 3 6 3 In selecting an objective test, Parliament appears to have accepted that transactions do serve objective purposes, and that reasonable taxpayers could recognize the purposes of tax-reducing transactions without being constrained by the motives and intentions of the particular taxpayer who arranged them. The second reason why focusing on taxpayers' reasons for arranging their transactions is not the appropriate approach to identifying the purpose of their transactions relates to the second branch of the definition of avoidance transaction. In paragraph 245(3)(b), Parliament dealt with series of transactions and applied an objective purpose test to each step in a series of transactions in the following terms: ) [an avoidance transaction means any transaction] that is part of a series of transactions, which series, but for [section 245], would result, directly or indirectly, in a tax benefit, unless the transaction may reasonably be considered to have been undertaken or arranged primarily for bona fide purposes other than to obtain the tax benefit. Where a series of transactions results in a tax benefit, paragraph 245(3)(b) mandates an examination of each transaction or step in the series to determine whether it was primarily undertaken for a bona fide purpose other than obtaining the tax benefit. If an individual step in the series fails the primary purpose test, it is an avoidance transaction and the tax consequences of that step must be redetermined. In enacting this provision, Parliament recognized that taxpayers pursuing economic or business objectives wi l l often insert Arnold, "Canadian General Anti-Avoidance Rule", supra note 354 at 231-32. Canada Trustco, supra note 80 at 5528, paras. 28-29. 130 additional steps into their transactions for the sole or primary purpose of reducing the tax ultimately payable. Parliament thus enacted a statutory step transaction doctrine. 3 6 4 It appears that paragraph 245(3)(b) was intended to function similarly to the U K and American step transaction doctrines, which deal specifically with circumstances in which taxpayers pursuing otherwise valid non-tax objectives insert tax-motivated transactions into a series of transactions for tax avoidance purposes. 3 6 5 Given that Parliament intended the test for an avoidance transaction to be an objective test, and that Parliament provided for a step transaction approach that applies the purpose test to each individual transaction in a series of transactions, the logical conclusion is that the appropriate way of reading the test for an avoidance transaction is that the purpose of a transaction is distinct from a taxpayer's reasons for arranging it. Where it is reasonable to consider that those purposes are tax benefits, transactions wi l l be avoidance transactions so long as the other requirement of subsection 245(3) (the primary purpose requirement) is met. i i . Primarily for Bona Fide Non-Tax Purposes The second question that arises when applying the purpose test in subsection 245(3) is how to identify legitimate purposes not caught by the G A A R . Unlike the traditional business purpose test, which is narrowly confined to transactions undertaken in a business context, subsection 245(3) refers to any "bona fide purpose other than to obtain the tax 3 6 4 Such an interpretation is supported by the explanatory notes released by the Department of Finance on the introduction of the GAAR amendments. The notes explaining paragraph 245(3)(b) speak of a "step transaction" as one transaction that is part of a series of transactions that is not inserted and carried out primarily for bona fide non-tax purposes. Refer to Department of Finance, Technical Notes, 17th ed., ed. David M. Sherman (Toronto: Thomson Canada Limited, 2005) at 2265. 3 6 5 For instance, the transactions in Gregory v. Helvering, supra note 101, and Furniss v. Dawson, supra note 180, are examples of inserting tax-motivated steps into a series of transactions ultimately undertaken in pursuit of a non-tax commercial objective of selling shares. 131 benefit." In enacting that test, Parliament intended to create a broader non-tax purpose test than the narrow business purpose test articulated in Gregory v. Helvering.366 The test laid down in subsection 245(3) is more like the broad non-tax purpose test articulated in Knetsch.361 Parliament recognized that, in a modern economy, taxpayers w i l l engage in transactions for a variety of non-business reasons, such as investment, family and estate planning reasons, and therefore enacted a broader test that would not automatically catch all transactions entered into outside a business context. 3 6 8 Therefore, the majority of personal, family, and investment transactions that have primary bona fide non-tax purposes should not be considered avoidance transactions even though they are not undertaken with a business purpose in the strict sense. 3 6 9 Nonetheless, although avoidance transactions are defined in terms of broad non-tax purposes, the test for an avoidance transaction is functionally the 170 same as a business purpose test. In addition to referring to non-tax purposes generally as opposed to a business purpose specifically, subsection 245(3) requires that those purposes be "bona fide." It is not clear what the bona fide requirement means, or how the bona fides of a transaction's purpose 366 Ibid. In Canada Truscto, supra note 80 at 5527-28, para. 21, the Supreme Court of Canada confirmed that the test is a broad "non-tax purpose" test. 367 Supra note 167. The broader non-tax purpose test developed in Knetsch is reviewed in Chapter 4 under the heading "The Business Purpose Test." 3 6 8 Arnold & Wilson, "Part II", supra note 96 at 1155. Also the Explanatory Notes confirm that subsection 245(3) refers to "bona fide purposes other than to obtain the tax benefit" rather than to "bona fide business purposes" precisely because of the concern over importing a narrowly-construed business purpose test. Thus, the GAAR is Parliament's response to the Supreme Court of Canada's decision to reject a business purpose test on the grounds that a narrow business purpose test would be problematic in a sophisticated revenue statute that included numerous tax incentive provisions. 3 6 9 This point is also confirmed in the explanatory notes. Technical Notes, supra note 364. 3 7 0 See Dodge, supra note 11 at 19. Further, Arnold, "Business Purpose Test", supra note 100 at 10:12, argues that there is no longer any meaningful distinction between a "business purpose test" and a "non-tax purpose test," since the test has naturally evolved to match the development of increasingly sophisticated fiscal legislation and economic activity. The phrase "business purpose test" may now be nothing more than shorthand for a broad non-tax purpose test a la Knetsch, supra note 167. 132 should be determined. 3 7 1 No cases have seriously considered whether the inclusion of the phrase "bona fide" has a significant effect on the operation of subsection 245(3). Arnold and Wilson suggest that the bona fide component "simply means that the relevant purposes must be genuine." 3 7 2 Their conclusion seems reasonable, given that subsection 245(3) refers to whether the transaction may reasonably be considered to have been undertaken or arranged primarily for non-tax purposes, creating an objective test asking if it is reasonable to consider a transaction to have been undertaken for a non-tax purpose. Obviously, whether a taxpayer had a genuine non-tax purpose in mind is related to the inquiry into whether it is objectively reasonable to consider that the transaction was undertaken for non-tax purposes. Finally, a transaction resulting in a tax benefit wi l l escape being an avoidance transaction under subsection 245(3) of the Act i f it is undertaken "primarily" for bona fide non-tax purposes. When transactions clearly serve no purpose other than obtaining tax reductions, characterizing their primary purpose is relatively straightforward. This inquiry may prove difficult, however, in complicated cases where transactions serve more than one purpose. Where a transaction serves multiple tax and non-tax purposes, identifying a "primary" purpose could be quite challenging. However, a primary purpose must be identified because transactions undertaken primarily for purposes of obtaining tax benefits are avoidance transactions even if they also serve other purposes. In fact, it is not even clear what "primary" means. On the one hand, "primary" could be interpreted in a relative sense, meaning simply that the tax purpose is the single most important, or chief, purpose among all of the competing purposes. For instance, the Concise Oxford Dictionary includes the following meanings for primary: "of the first important, 3 7 1 As Couzin observes, the words "bona fide" reflect language used in some business purpose cases, but it is not clear what they add to the definition of "avoidance transaction." Couzin, supra note 359 at 4:7. 3 7 2 Arnold & Wilson, "Part II", supra note 96 at 1156. chief, of the first rank in a series." On the other hand, "primary" could be interpreted in a more absolute sense, meaning that the relevant purpose must be closer to a "sole" or "dominant" purpose, or that non-tax purposes must outweigh tax purposes (i.e., they must exceed a 50% weighting). 3 7 4 The fact that subsection 245(3) itself refers to "bona fide purposes" in the plural suggests that Parliament intended all of a transaction's non-tax purposes to be aggregated, rather than viewed individually, in order to assess whether the transaction was primarily undertaken for non-tax purposes. Theoretically, the interpretation given to "primary" could have a significant effect on the application of the G A A R whenever a transaction has multiple purposes but no one purpose could be described as dominant or as exceeding 50%. For example, assuming a tax-deferred share exchange transaction had three purposes - tax deferral, estate planning, and investment - and assuming that it was possible to determine that the transaction was 40% tax-motivated and 30% motivated by each of the estate planning and investment purposes, would the transaction be an avoidance transaction? If the primary purpose test requires simply that tax deferral be the most important single purpose, then the transaction would indeed be an avoidance transaction. If, on the other hand, the primary purpose test requires that the tax deferral be the sole or dominant purpose, or at least that tax purposes outweigh all other purposes, it would be harder to argue that the transaction is an avoidance transaction, because the combined non-tax purposes outweigh the tax purpose. There has not been a great deal of discussion or debate over the precise meaning of the primary purpose test. The courts have recognized that subsection 245(3) requires them to 373 Concise Oxford Dictionary, 9* ed. (Oxford: Clarendon Press, 1995). 3 7 4 Arnold and Wilson, "Part II", supra note 96 at 1161, observe that a "sole" or "dominant" purpose interpretation would be more consistent with the business purpose test applied in the United States and the step transaction test applied in the UK (at least before the developments in McGuckian, MacNiven, and Barclays Mercantile described in Chapter 4). 134 weigh the evidence and determine whether it is reasonable to conclude that the transaction in issue was undertaken primarily for non-tax purposes. 3 7 5 However, there has been little judicial discussion of the meaning of "primarily" beyond acknowledging the requirement to 376 evaluate the relative importance of the various "driving forces" behind transactions. In part, this might be because, in most cases brought forward by the tax authorities, the facts have made identifying the primary purpose of the transaction relatively easy. In addition, the Department of Finance issued Explanatory Notes indicating that the intention of the primary purpose test is to exclude transactions that incidentally result in a tax benefit or transactions for which tax considerations were a significant, but not the primary, purpose for carrying out 177 the transaction. This issue arises whenever purpose-oriented tests are used to identify tax avoidance. As discussed in Chapter 3, purpose-based approaches inevitably raise the question of to what degree tax must be a motivating factor before transactions wi l l be considered abusive. Selecting such a degree, or threshold, is a key step in identifying abusive tax avoidance because that threshold involves balancing taxpayers' need to pursue economic and other non-tax goals in a tax-efficient way, taking advantage of acceptable opportunities for tax mitigation, with a societal need to prevent taxpayers from cloaking tax-driven transactions with minimal or ancillary commercial considerations simply for tax avoidance purposes. Parliament has chosen to set the requirement at the level of "primary," which appears to be a relatively high threshold. In light of the need to distinguish between legitimate tax mitigation and tax avoidance, I believe that the primary purpose test ought to be interpreted to collectively weigh all of a transaction's non-tax purposes against its tax purposes, rather than 375 Canada Trustco, supra note 80 at 5528, para. 29. 376 Ibid, at 5528, para. 28. 377 Technical Notes, supra note 364 at 2265. 135 by identifying each individual purpose in isolation. Where a transaction's tax purposes outweigh the other non-tax purposes, it should be an avoidance transaction. i i i . Series of Transactions One final issue that arises in connection with subsection 245(3) is how to interpret the phrase "series of transactions." As discussed above, paragraph 245(3)(b) is intended to create a statutory step transaction doctrine applying the objective primary purpose test to each step in a series of transactions. 3 7 8 Each step must have been undertaken primarily for non-tax purposes; i f any step fails this test, its tax consequences wi l l be redetermined in order to deny the tax benefit. It is important to note that any transactions caught as avoidance transactions would still be legally effective. Only their tax consequences would be redetermined under the G A A R . As a result, the series of transactions would still reach its original economic and legal result, but would not enjoy the beneficial tax consequences that would otherwise have resulted from the intermediate step. For example, in a case like 1*7Q Gregory v. Helvering, the incorporation of an intermediate holding company and transfers of shares would fail the primary purpose test and would be avoidance transactions. As long as the transactions were not saved by the exception in subsection 245(4), the G A A R would "l o n deny those transactions tax-exempt status. However, the incorporation of the intermediate holding company, the transfers of shares, and its winding-up would still be effective. As a result, the taxpayer would still hold the shares, although there would be tax consequences on 3 7 8 Arnold & Wilson, "Part II", supra note 96 at 1161. 379 Supra note 101. The facts were described in Chapter 4 under the heading "The Business Purpose Test." 3 8 0 This example assumes, of course, that had Mrs. Gregory carried out her transactions in Canada, identical tax-exempt rollover provisions would apply. 136 the transfer of those shares to the intermediate company, and she would have a taxable dividend resulting from the receipt of those shares on the winding-up of that company. In addition, because paragraph 245(3)(b) merely requires that the series of transactions result in "a" tax benefit and is not specific about the identity of the taxpayer receiving the benefit, a transaction wi l l be an avoidance transaction even if the resulting tax benefit is not received by a party to the transaction but by another taxpayer who is party to a different transaction later in the series. This rule therefore permits the G A A R to apply to tax benefits realized by taxpayers who enter a series of transactions merely to complete the ultimate series, and who are not party to the original steps that lay the foundation for realizing the tax benefit. Such a situation occurred in Matthew, where the initial transactions undertaken by Standard Trust and a non-arm's length partnership were avoidance transactions even though the tax benefits were not realized by them, but by investors who 381 purchased units in a different partnership later in the series of events. The most significant issue that arises in connection with the step transaction test of subsection 245(3) is determining exactly what constitutes a series and whether a transaction giving rise to a tax benefit is sufficiently connected to a series of other transactions to be considered part of that series. Chapter 4 identified three different methods of formulating this nexus in the United States and U K case law: the binding commitment, mutual independence, and end results tests. 3 8 2 These three versions of the step transaction doctrine differ in terms of the degree of pre-ordination, or commitment, that is required to connect a later transaction to a prior transaction. In Canada, the courts have adopted the most 3 8 1 Indeed, this sort of situation often occurs when preliminary transactions are undertaken to "package" losses or capital cost allowance for marketing to other taxpayers who can access them through partnerships and deduct them from other income. In those cases, both the preliminary "packaging" transactions and the ultimate purchase of partnership units will be avoidance transactions because they are part of the same series. See the discussion in Chapter 4 under the heading "The Step Transaction Doctrine." 137 restrictive common law formulation, the binding commitment test, as the meaning of a "series of transactions" at common l a w . 3 8 3 Thus, in Canada, a transaction wi l l be part of a series of transactions at common law only if it can be said, upon the occurrence of the first transaction, that it is pre-ordained that the subsequent transactions wi l l occur. However, in its most recent comments on this issue, the Supreme Court of Canada explained pre-ordination as meaning that "there is no practical likelihood that the pre-planned events would not take place in the order ordained," as opposed to meaning that there is a binding commitment that the subsequent steps would take p lace . 3 8 5 The Canadian courts' relatively restrictive approach to identifying series of transactions (at least compared to United States courts) is not particularly significant for the operation of the G A A R , however, because of subsection 248(10) of the Act. That provision is a broad deeming provision that captures many more transactions than the common law "no practical likelihood" nexus does. Subsection 248(10) deems a series to include "any related transactions.. .completed in contemplation of the series." Such a broadly-worded provision connects, as part of a series, any transaction completed in contemplation of the series, regardless whether it was at all pre-ordained that the subsequent transaction would occur. In fact, under subsection 248(10) it is not even necessary that the parties to a subsequent transaction know of the preceding transactions and take them into account in completing the subsequent transaction. A l l that is necessary to connect the subsequent 383 Canada Trustco, supra note 80 at 5528, para. 25. Also see OSFC Holdings, supra note 346 at 5479, para. 24. 384 Canada Trustco, ibid, at 5528, para. 25. 385 Ibid, at 5528, para. 25. This formulation is drawn from the House of Lords' decision in Craven v. White, supra note 200, which, as discussed in Chapter 4 under the heading "The Step Transaction Doctrine in Canadian Law," sets a higher threshold than the end result test, but is not necessarily the same thing as the binding commitment test originally set out in Ramsay, supra note 179. The Court purported to agree with the approach taken by the Federal Court of Appeal in OSFC Holdings, supra note 346 at 5479, para. 24, which had been the leading Canadian GAAR case before Canada Trustco and Matthew. 386 Income Tax Act, supra note 7, ss. 248(10). 138 transaction to a prior series is that the subsequent transaction relate to and be undertaken because of the prior series. 3 8 7 Thus, speculative transactions, such as i f the vendors in TOO Craven v. White had transferred those shares to an offshore holding company even before they began negotiating with a potential purchaser, merely in anticipation that they might choose to sell their shares some day, would be connected with subsequent transactions even if the subsequent transactions might not have been predictable, let alone pre-ordained, at the time of the initial transactions. Furthermore, given that the words " in contemplation o f in paragraph 248(10) have been interpreted so broadly, it is possible that transactions that took IRQ place before an avoidance transaction can be connected into a subsequent series. The series issue arose in Matthew because, at the time that the liquidator first transferred Standard Trust's mortgage portfolio to a non-arm's length partnership, the subsequent sale of the partnership interest to another partnership and the acquisition of units in that second partnership by outside investors like Matthew were not pre-ordained in the common law sense. Although it was predictable, i f not intended, that the 99% interest in Partnership A would be sold to other parties who could use the accrued tax losses to offset other income, no such purchasers and investors were known to Standard Trust's liquidator, who was merely rearranging Standard Trust's assets into a marketable form in the hope of finding a purchaser. Given the absence of certainty about the purchasers, it could not be said that there was "no practical likelihood" that those subsequent transactions would not take place. Unless subsection 248(10) connected the subsequent sale of the 99% interest in 3 8 7 In OSFC Holdings, supra note 346 at 5480-81, para. 36, the Federal Court of Appeal decided that the words "in contemplation o f in subsection 248(10) required that the parties to the subsequent transaction know of the series "such that it could be said that they took it into account when deciding to complete the transaction." In Canada Trustco, supra note 80 at 5528, para. 26, the Supreme Court of Canada explicitly overruled the Federal Court of Appeal and decided that subsection 248(10) does not require any actual knowledge, merely a relationship between the transaction and the series. 388 Supra note 200. 389 Canada Trustco, supra note 80 at 5528, para. 26. 139 Partnership A to Partnership B and the sale of Partnership B units to outside investors into the series, then, those subsequent transactions would not be part of the series for purposes of applying the G A A R and the tax benefits realized by the investors might not be denied. Based on the Supreme Court of Canada's reading of " in contemplation o f in subsection 248(10), however, there is no doubt that the subsequent steps in the Matthew case are part of a series of transactions for the purposes of the G A A R . 3 9 0 E. The Exception for Misuse, Abuse and Frustration The third stage of the G A A R analysis is the saving or exemption clause in subsection 245(4). Under that provision, the G A A R applies only to transactions that may reasonably be considered to result, directly or indirectly, in a misuse of statutory or regulatory provisions, or in an abuse having regard to those provisions read as a whole. Even i f a transaction qualifies as an avoidance transaction under subsection 245(3), therefore, it w i l l not be subject to the G A A R unless it also reasonably results in a misuse or an abuse. Parliament intends not to deny effect to all tax-reducing transactions entered into primarily for tax purposes, but only to those that, for policy reasons, are abusive. Subsection 245(4) is at the heart of the determination of whether transactions are abusive. Unfortunately, however, the Act is not at all clear as to what constitutes a misuse or abuse. There has been much debate and 3 9 0 A similar issue arose in OSFC Holdings, supra note 346. OSFC initially purchased the 99% interest in Partnership A and later re-sold that interest to Partnership B, becoming a 24% owner of Partnership B. In OSFC Holdings, the Court did not need to consider whether the subsequent sale of Partnership B units to outside investors like Matthew were avoidance transactions and part of a series, since that case concerned only OSFC's claim for a partnership loss as a 24% partner in Partnership B. 3 9 1 The Department of Finance's explanatory notes shed a little light on this issue, although they are not completely clear themselves. The notes suggest that the GAAR is intended to apply to transactions carried out primarily to obtain tax benefits not intended by the specific provisions relied upon and by the Act as a whole. The notes also suggest that, when a transaction relies on specific provisions of the Act to achieve a tax benefit, 140 commentary on this question, both in the literature and in the case law, since the G A A R was introduced. Much of the debate has revolved around whether misuse and abuse are the same or different concepts, whether they are terms of art or simply refer to results that are not intended by the Act, and whether the provision does any more than simply reinforce the modern rule of interpretation originally articulated in Stubart that requires the Act to be read in light of its object and spirit and not merely its literal, textual meaning. 3 9 2 Unti l October 2005, OSFC Holdings393 was the leading G A A R decision, establishing that misuse and abuse were different concepts and that determining whether a transaction resulted in either a misuse or an abuse involved, first, identifying a policy underlying the relevant provisions at issue; and, second, assessing whether the transaction resulted in a misuse or abuse having regard to that policy. In OSFC Holdings, the Court actually referred to "policy" in the context of describing the approach to identifying a misuse or abuse as a "purposive, object and spirit, scheme or policy" approach to identifying a misuse or abuse. Taken in context, the Court did not mean that the results of a transaction must be found to offend some extra-statutory Parliamentary policy before the G A A R would apply. Rather, the the GAAR is intended to mandate an object and spirit reading of those provisions, and of the Act as a whole, to determine whether the tax benefits are intended by Parliament. However, the notes are confusing in that they suggest that a misuse of provisions is conceptually distinct from an abuse of the Act as a whole, and that subsection 245(4) draws on the abuse of rights doctrine discussed in Chapter 4. The notes do not offer any detail as to what the GAAR shares in common with the abuse of rights doctrine, other than the obvious similarity that taxpayers should not be entitled to deliberately take advantage of provisions in ways that they were not intended to be used to achieve tax benefits that Parliament did not intend them to achieve. Technical Notes, supra note 364 at 2266. 3 9 2 For instance, Arnold & Wilson, "Part II", supra note 96 at 1165-66, argue that the misuse and abuse requirement must mean more than simply a restatement of the object and spirit approach. In their view, Stubart mandates an object and spirit interpretation of all provisions, and thus if a taxpayer was entitled to a tax benefit on an object and spirit interpretation of the relevant provisions before considering the GAAR, the subsection 245(4) exception would be meaningless unless it required more than simply another review of the object and spirit of the provisions. More recently, however, Arnold argues for identifying a misuse or abuse based on an object and spirit- or statutory scheme-based reading of the Act. "Long, Slow Steady Demise", supra note 58 at 496-504. For another view favouring an emphasis on object and spirit, refer to Dodge, supra note 11, who argues at 20-21 that subsection 245(4) should involve an analysis of the object and spirit of the provisions relied upon by the taxpayer to justify the tax benefit. 393 Supra note 342. 141 Court intended that the "policy" of the provisions could be identified by reviewing their object and spirit, or scheme. The Court likely meant that, in order to identify a misuse or abuse, the legislative provisions used by a taxpayer should be given a purposive, and not literal, reading, and the question should be whether the tax benefit claimed by the taxpayer is one that Parliament intended them to receive. Unfortunately, later cases fixated on the word "policy," and overlooked the Court's references to object and spirit or scheme, with the result that the courts became focused on why Parliament would want certain tax consequences, as opposed to merely determining which tax consequences Parliament intended to apply to the outcome of taxpayers' transactions. The former question calls for a review of extra-statutory material and tax policy questions, while the latter question wi l l often be answered by considering the object and spirit, or purpose, of the provisions in question and their place in the statutory scheme as a whole . 3 9 4 In addition, OSFC Holdings indicated that courts should be cautious about denying tax benefits resulting from compliance with technical provisions of the Act and should do so only when the relevant policy of the provisions was clear and unambiguous. Generally, such a cautious approach is appropriate, since the G A A R amounts to a kind of extraordinary statutory override whereby one statutory provision nullifies the effect of other, relatively clear, statutory provisions with which a taxpayer has complied. However, the requirement for a clear and unambiguous policy resulted in the courts placing an increasing burden on the tax authorities to advance a clear and unambiguous policy before the G A A R could apply, and 3 9 4 Arnold argues convincingly that the courts erred in their approach to identifying a misuse and abuse by fixating on policy without grounding that inquiry in the object and spirit, or scheme, of the Act itself. "Long, Slow Steady Demise," supra note 58 at 498-504. Arnold also observes that, perhaps unwittingly, the Courts' fixation on underlying Parliamentary policies resulted in an interpretation that was unduly favourable to taxpayers, given the difficulty that the tax authorities face in articulating (and proving) a clear and unambiguous policy underlying complex provisions of the Act, which often serve multiple competing Parliamentary purposes at one time. 142 in some cases the courts appeared to decline to apply the G A A R almost by default because the tax authorities had not articulated a sufficiently clear and unambiguous underlying po l i cy . 3 9 5 However, in Canada Trustco, the Supreme Court of Canada largely rejected the OSFC Holdings approach to identifying a misuse and abuse and articulated a new interpretive approach to subsection 245(4). The Court held, first, that there is no difference between a misuse of a statutory provision and an abuse of the statute as a whole. The Court did not accept that taxpayers could abuse the Act as a whole without misusing any of its provisions. 3 9 6 In the Court's view, the G A A R is intended to distinguish between abusive tax avoidance and legitimate tax mitigation, and questions of whether abusive tax avoidance is a misuse, an abuse, or something else are not really important. In addition, the Court rejected the two-stage approach articulated in OSFC Holdings that required identifying a policy underlying the provisions of the Act in order to determine whether transactions result in a misuse or abuse. In the Supreme Court's view, the provisions of the Act must be interpreted " in their contextual framework, so that the Act functions as a coherent whole." Restating the modern rule of interpretation, the Court held that there is "but one principle of interpretation: to determine the intent of the legislator having regard to the text, its context, and other indicators of legislative purpose." Therefore, the Court held, the search for "policy" urged in OSFC Holdings, which in subsequent cases had become an almost extra-3 9 5 For instance, Hill, supra note 56. For vigorous criticism of the decision in Hill, refer to Arnold, "Long, Slow Steady Demise", supra note 58 at 500-01. The facts of Hill are reviewed in Chapter 3 under the heading "The Nature of the Tax System." Also refer to the subsequent discussion of how the Tax Court erroneously concluded that there was no discernable Parliamentary policy in Hill. 396 Canada Trustco, supra note 80 at 5530, para. 39. The Court was also influenced by the fact that the French version of subsection 245(4) refers merely to an "abus," not to a misuse and an abuse separately. The discrepancy between the English and French versions could not but influence the Court, given the requirement to ensure that equal effect is given to both official versions of Canadian statutes. 397 Ibid, at 5530, para. 39. 398 Ibid, at 5530, para. 40. statutory ideal against which to measure the appropriateness of an avoidance transaction, is not a distinct stage of the analysis. Rather, the search for the "policy" of the provisions is really a search for their underlying purpose, and is properly merged into "a unified textual, contextual and purposive approach to interpreting the specific provisions that give rise to the tax benefit." 3 9 9 The Court rejected the idea that the G A A R requires the Court to identify the underlying tax policy reasons for Parliament's decision to apply particular tax consequences to the outcomes of taxpayers' transactions. 4 0 0 The Court observed that to do so would change the role of judges from one of interpreting and applying the Act as it is written, in order to implement Parliament's intent, to one of deciding cases on the basis of appropriate tax policy outcomes formulated by judges themselves. 4 0 1 The Court was correct to make this observation. In previous cases, the courts had become obsessed with the "why" behind the structure of the Act, often overlooking clear legislative schemes from which a policy - in the sense of Parliament's intent - was readily discernable and should have been given effect through the G A A R . For instance, in Hill,402 the Tax Court failed to appreciate that Parliament's policy, or intent, was to treat compound and simple interest differently. Because the Tax Court could not understand why Parliament made that choice, it refused to apply the G A A R to a transaction that superficially complied with the technical requirements of the Act but resulted, in effect, in a taxpayer deducting compound interest on a payable basis. 4 0 3 To the extent that Canada Trustco prevents courts from embarking on a policy-399 Ibid, at 5530, para. 40. 4 0 0 Ibid, at 5530, para. 41. 401 Ibid, at 5530, para. 41. 4 0 2 Supra note 56. 4 0 3 Again, refer to the discussion of the facts of Hill and the relevant provisions in Chapter 3 under the heading "The Nature of the Tax System." 144 oriented inquiry as to why Parliament elected to apply certain tax consequences, and requires courts considering a misuse or abuse to identify the tax consequences that Parliament intended to result from a taxpayer's transaction - regardless why Parliament intended them -the Court's decision is a welcome clarification of how the G A A R should operate. 4 0 4 According to Canada Trustco, the ultimate question to be answered when applying subsection 245(4) is, "having regard to the text, context and purpose of the provisions on which the taxpayer relies, whether the transaction frustrates or defeats the object, spirit or purpose of those provisions." 4 0 5 Whether a transaction results in abusive tax avoidance is determined by "conducting a unified textual, contextual and purposive analysis of the provisions giving rise to the tax benefit in order to determine why they were put in place and why the benefit was conferred." 4 0 6 In the Court's view, abusive tax avoidance results when a taxpayer relies on specific legislative provisions to achieve a tax outcome that those provisions seek to prevent, when a transaction defeats the underlying rationale of the provisions relied upon, and where an arrangement circumvents the application of certain provisions in a manner that frustrates the object, spirit, or purpose of those provisions. 4 0 7 In other words, where a taxpayer relies on specific provisions of the Act to secure a tax benefit, the question to be answered is whether, considering all of the provisions of the Act as a 4 0 4 Unfortunately, although the Supreme Court of Canada rightly rejected a searching tax policy inquiry in favour of a focus on the scheme of the Act, its application of this approach in Canada Trustco is problematic because, as I will argue in the next Chapter, the Court failed to take a sufficiently instrumentalist view of tax legislation. Tax legislation is intended ultimately to apply tax consequences to underlying economic transactions and results, and in the Canada Trustco case the Court generally rejected the submission that underlying economic circumstances and outcomes are relevant for purposes of applying the GAAR. However, in my view, the GAAR serves as the "provision of last resort" to be used to ensure that the tax consequences mandated by law for taxpayers' transactions accord with those that Parliament would have intended as a matter of fiscal policy. 405 Canada Trustco, supra note 80 at 5531, para. 49. The Court adopted the phrase "frustration" as an appropriate analogue to replace "misuse" and "abuse." 4 d 6 Ibid, at 5533-34, para. 66. 407 Ibid, at 5530-31, para. 45. 145 harmonious whole, the tax benefit was intended to be available to that taxpayer in the circumstances. Notably, the Court observed that, according to the Explanatory Notes, the provisions of the Act are intended to apply to transactions with real economic substance and not merely to transactions intended to "exploit, misuse or frustrate the Act to avoid tax." 4 0 8 In doing so, the Court acknowledged that whether transactions have economic substance is sometimes relevant to whether they result in a misuse or abuse. However, the Court limited the relevance of economic substance to instances where, on an object and spirit reading of the relevant statutory provisions, it can be said that a particular tax benefit is available only if transactions have economic substance or "certain economic, commercial, family or other non-tax purpose[s]." 4 0 9 The Court rejected the assertion that, generally, transactions without economic or commercial purposes are abusive under subsection 245(4). In doing so, the Court effectively rejected the notion that the G A A R preserves any economic substance requirement underlying the provisions of the Act as a w h o l e . 4 1 0 Finally, the Court reiterated the importance of considerations of certainty and predictability, stating that Parliament accepts the Duke of Westminster principle that taxpayers are entitled to arrange their affairs so as to minimize taxes and that, in order to do so, tax law must be as predictable and reliable as possible. 4 1 1 In the Court's view, the G A A R 408 Ibid, at 5531, para. 49. 4 0 9 Ibid, at 5532, para. 58. 4 1 0 As I will argue in the final Chapter, the Court appears to have misunderstood the meaning of economic substance. In my view, the reference in the Explanatory Notes to "transactions with real economic substance" is a statement that tax law is intended to apply legal consequences on the basis of underlying economic activities and outcomes. When those activities and outcomes are not present, even if the legal relationships that trigger tax consequences have been created, Parliament would not have intended the same tax consequences to result. 4 1 1 The Court is not without some basis for doing so. The Explanatory Notes issued by the Department of Finance themselves recognize that "tax planning - arranging one's affairs to attract the least amount of tax - is a legitimate and accepted part of Canadian tax law." 146 should not be interpreted in a way that inhibits legitimate tax mitigation. For this reason, the Court held that the G A A R should be applied only i f it is not reasonable to conclude that a transaction does not result in a misuse or abuse having regard to the object and spirit of the relevant provisions of the Act. In other words, the Court enunciated a sort of default rule that gives taxpayers carrying out avoidance transactions the benefit of the doubt with respect to their tax benefits. 4 1 2 F. Conclusions In Chapter 3,1 identified three common approaches to identifying abusive tax avoidance: form-based, purpose-based, and policy-based approaches. Whichever approach is adopted, the challenge is to strike an appropriate balance between the need to control abusive tax avoidance and the desire to constrain legitimate tax mitigation as minimally as possible. Like the business purpose test and step transaction doctrine discussed in Chapter 4, the G A A R uses a purpose-based approach (the test for an avoidance transaction under subsection 245(3)) to identify tax avoidance. However, unlike the business purpose test and 4 1 2 It is quite likely that this rule may already have been superseded. At the time of the transactions in Canada Trustco, subsection 245(4) was phrased in a double negative: "subsection (2) does not apply to a transaction where it may reasonably be considered that the transaction would not result directly in a misuse.. .or an abuse..." Subsection 245(4) therefore permitted a taxpayer to escape the GAAR if it could reasonably be considered that their transaction was not abusive. At paragraph 62, the Court seemed to elevate the double-negative wording of subsection 245(4) into a presumption that "a tax benefit that would be conferred by the plain words of the Act is not abusive." However, subsection 245(4) has since been amended to provide that the GAAR applies only if it may reasonably be considered that the transaction results in a misuse or abuse. Some authors, such as Duff, suggest that the new wording is a lower threshold for the application of the GAAR. It may be that the GAAR is more likely to apply if all that is necessary is that it be reasonable to consider a transaction abusive as compared to it being reasonable to consider that a transaction is not abusive. Then again, these differences may simply be semantics. Whether the amendment has any effect on the development of the GAAR in Canadian law remains to be seen. See David Duff, "The Supreme Court of Canada and the General Anti-Avoidance Rule: Canada Trustco and Matthew," prepared for the University of Toronto GAAR Symposium November 18, 2005, at 44-45. 147 step transaction doctrine, the G A A R purpose test merely identifies transactions that could be tax avoidance. The G A A R then employs a policy-oriented approach (the misuse and abuse, or frustration, test in subsection 245(4)) to determine which transactions are nonetheless permissible even though they fail the non-tax purpose test. Parliament's reason for supplementing a purpose-based approach with a policy-oriented safeguard is that a purely purpose-focused test is potentially too broad, catching some forms of tax mitigation that ought to be legitimate, such as taking advantage of tax incentive provisions. The "misuse and abuse" exclusion permits transactions that have a primary, or even sole, tax purpose to nevertheless be respected for tax purposes i f their results accord with the object and spirit of the Act (or, alternatively, so long as their results do not frustrate the intent of Parliament). The G A A R therefore represents Parliament's balance between controlling abusive tax avoidance and permitting legitimate tax mitigation. This thesis now turns to a review of whether the G A A R , as drafted by Parliament and as applied by the courts, accords with good tax policy. Specifically, in the context of two specific tax avoidance scenarios - a circular financing sale-leaseback and surplus stripping - may the G A A R ensure that normative ideal tax outcomes are reached, and, i f it has not done so in the decided cases to date, why and how has the law failed to do so? 148 VI. APPLYING THE GAAR TO ENSURE APPROPRIATE OUTCOMES FROM A TAX POLICY PERSPECTIVE In Chapter 3,1 suggested that an appropriate framework for understanding abusive tax avoidance is that tax avoidance results when, because tax legislation applies only to the legal forms and relationships created by taxpayers' transactions, the legislation fails to impose the tax consequences that, from a tax policy perspective, would have been intended to apply to their underlying economic outcomes. I termed those tax consequences the "normative ideal" tax consequences. In addition, abusive tax avoidance involves transactions engaged in by taxpayers for the sole or primary purpose of obtaining such a tax benefit. Recognizing that taxpayers are entitled to engage in legitimate tax mitigation, however, transactions should not be considered abusive where a tax benefit is anticipated by Parliament and encouraged or provided for in the very scheme of the Act itself. In those circumstances, the tax benefit is the normative ideal tax outcome. Because Canadian courts have declined, at virtually every opportunity, to adopt judicial general anti-avoidance principles, such as a business purpose test, step transaction doctrine, or economic substance doctrine, the G A A R is the only meaningful anti-avoidance rule in Canadian tax law that could prevent taxpayers who successfully navigate the maze of specific statutory rules from obtaining tax benefits in a manner that frustrates the intention of Parliament. Therefore, having reviewed the structure of the G A A R and the manner in which Canadian courts have interpreted its components, the next question is whether the G A A R , as it is designed and as the courts have interpreted it, properly identifies abusive tax avoidance and preserves normative ideal tax outcomes. 149 I believe that an appropriate approach to this analysis is through case studies examining tax avoidance scenarios that result in abusive tax avoidance as defined in Chapter 3. Two such scenarios are circular financing sale-leasebacks and surplus stripping. The logical first step in this analysis is to review the issues arising in those scenarios and why the Canadian courts have or have not applied the G A A R . Next, the inquiry should identify the normative ideal tax consequences that Parliament would have intended to apply to the outcome of those transactions, which wi l l demonstrate why those transactions are abusive. Finally, this Chapter w i l l explain why the G A A R should apply in those cases, and how the courts have failed to appreciate the fundamentally instrumental nature of tax law and to interpret the G A A R as vigorously as necessary in order to give effect to Parliament ultimate intent. Ultimately, I conclude that the G A A R is not inherently flawed. It represents a reasonable balance between the need to identify and prevent abusive tax avoidance and the need to preserve legitimate tax mitigation opportunities. To strike that balance, the G A A R employs an objective primary purpose test to identify avoidance transactions and a policy-based exception for transactions that do not misuse, abuse or frustrate the Act when read in accordance with object and spirit (a term loosely synonymous with Parliamentary intent). The Act is, first and foremost, a policy instrument intended to implement Parliament's underlying economic and fiscal policies through the application of specific tax consequences based primarily on taxpayers' underlying economic circumstances. However, Canadian courts have generally failed to appreciate the object and spirit of the Act as a whole. Therefore, Canadian courts have not applied the G A A R in the manner in which it should be 150 applied, and they have permitted abusive tax avoidance to continue and unacceptable tax policy outcomes to occur. A. A Circular Financing Sale-Leaseback: Canada Trustco Mortgage Company i . The Facts and Judgments The facts of Canada Trustco were outlined in the previous Chapter. Canada Trustco wanted to shelter approximately $50 mill ion of taxable leasing income, so it paid a corporate arranger to locate a sale-leaseback transaction that would entitle it to deduct capital cost allowance against that income. 4 1 3 The sale-leaseback used a combination of borrowed money and Canada Trustco's own funds to purchase trailers from T L I , a U S company. 4 1 4 Canada Trustco then leased the trailers to M A I L , an offshore intermediary which in turn leased the trailers back to T L I . T L I immediately prepaid the leaseback in full to M A I L , and M A I L deposited the same amount as Canada Trustco had borrowed with the bank that had originally loaned it, and the remainder in a secure bond pledged to Canada Trustco as security for M A I L ' S remaining lease obligations. Canada Trustco then assigned the deposited funds to the bank and directed the bank to use those funds to offset its periodic loan repayments as they came due. As the new owner of the trailers, Canada Trustco deducted capital cost allowance against other leasing income. 4 1 5 4 1 3 See the discussion in Chapter 5, supra note 341, for a description of how sale-leasebacks operate. 4 1 4 The fact that the assets purchased were trailers is significant because trailers are on the list of property exempt from specified leasing property rules that would otherwise restrict capital cost allowance. 4 1 5 Under paragraphs 18(l)(b) and 20(1 )(a) of the Act, and Regulation 1100 and Schedule II to the Regulations, capital cost allowance is claimed on trailers at the rate of 30%. 151 The transaction was structured so that Canada Trustco assumed no economic risks in acquiring the trailers. There was no risk of defaulting on its bank debt because the borrowed money was immediately cycled back to the bank as part of the transaction. There was no risk of losing its own funds through a default by M A I L because those funds were immediately invested in a secure instrument pledged back to Canada Trustco. From Canada Trustco's commercial perspective, the transaction was a gem because it acquired leasing assets, thus accessing capital cost allowance, without incurring any real economic cost or even risk. Because it used primarily borrowed money that was immediately returned to the lender, in economic terms Canada Trustco laid out no more than $25 mill ion to acquire the trailers and capital cost allowance deductions worth substantially more than that. 4 1 6 However, $19 mil l ion of that $25 mill ion was immediately beneficially returned to Canada Trustco before the transaction even completed, and would grow to $33 mill ion by maturity. The bond thus gave Canada Trustco an $8 mill ion return on its initial $25 mill ion investment in addition to the tax benefits. The tax planning behind this transaction was so extensive that it minimized or eliminated any of the tax consequences that would normally arise on a prepaid sale-leaseback. What emerges, in reality, is a complicated structure designed to manufacture a tax benefit without any change in the underlying economic circumstances of the parties. For instance, an immediate prepayment of T L I ' s lease payments directly back to Canada Trustco would have been taxable leasing income to Canada Trustco. 4 1 7 Therefore, the parties inserted the offshore intermediary M A I L so that an immediate prepayment could be made 4 1 6 In addition, Canada Trustco paid a total of $2.34 million in arrangement fees and transaction costs. Canada Trustco added this amount to its cost of the trailers for capital cost allowance purposes. 4 1 7 Indeed, an immediate prepayment in full might have amounted to a disposition of the assets back to TLI, thus depriving Canada Trustco of any capital cost allowance at all. 152 without creating tax consequences for Canada Trustco. 4 1 8 In addition, simply depositing the funds with the bank and executing a direction to offset those funds against Canada Trustco's monthly loan payments, instead of fully repaying the loan up front, permitted Canada Trustco to deduct ongoing interest expenses as a way of sheltering the notional leasing income received from M A I L . Furthermore, the transaction permitted the U S vendor to pocket $3.6 mil l ion representing the difference between the original sale price and its total lease repayment, and then to walk away from the transaction with no continuing obligations while still holding the trailers 4 1 9 Finally, the $19 mil l ion bond returning Canada Trustco's own up front investment was held by another offshore entity rather than by Canada Trustco directly so that any interest income on the bond would not be taxable to Canada Trustco as it accrued. Even the choice of assets used in the sale-leaseback was tax-motivated. Canada Trustco chose a transaction using trailers because trailers are exempt from specific regulations aimed at restricting capital cost allowance in respect of leasing property. Generally, leasing is an attractive financing alternative for cash-strapped taxpayers. They can sell business assets to cash-rich taxpayers who are interested in holding legal title to the assets in order to claim capital cost allowance against other income, but who do not need the assets themselves and are wil l ing to lease the assets back to the original vendors for continued use. In addition, capital cost allowance rates are typically more generous than the actual depreciation in the first years of an asset's lifetime, providing an added incentive to cash-rich taxpayers to engage in lease financing. Lease financing proved so popular that, in 4 1 8 In addition, the bank probably much preferred to finance a transaction that immediately returned the principal to the bank's control than a normal transaction where the bank would at best take security over the trailers. Furthermore, the bank controlled the offshore intermediary MAIL, thus ensuring that there was absolutely no risk to its own capital in this transaction. 4 1 9 TLI continued to hold legal title to the trailers under a trust agreement, with Canada Trustco, and even claimed its own depreciation expense in respect of the trailers for US tax purposes. 153 the 1976 Budget, Parliament introduced rules (known as the "specified leasing property rules") to restrict capital cost allowance in respect of leasing properties in order to protect the tax base from erosion occurring through proliferating lease financing schemes. 4 2 0 Later, additional regulations were enacted to exempt certain assets from the specified leasing property rules. The rationale for exempting certain assets was that where the capital cost allowance rates for assets reasonably approximated actual depreciation, those assets offered no incentive to engage in lease-financing solely to obtain tax reductions resulting from huge up-front deductions. 4 2 1 After purchasing the trailers, Canada Trustco claimed capital cost allowance on a capital cost of $122.34 mill ion. That cost figure was based on the legal form of the transaction. Legally, Canada Trustco had laid out $120 mil l ion for the trailers and $2.34 mill ion for arrangement fees and transaction costs under valid, binding agreements that created real, enforceable legal relationships. The legal cost is not affected by the fact that all money laid out was guaranteed to be returned to its origins. Because trailers were on the list of property exempt from the specified leasing property rules, there would not be any concern about the restriction of that capital cost allowance. Legally, therefore, Canada Trustco had complied with all of the requirements of the Act and would be entitled to deduct capital cost allowance unless the G A A R applies to the transaction. The Crown sought to apply the G A A R and to deny the capital cost allowance claimed. The Crown argued, and all levels of court agreed, that the transaction resulted in a tax benefit under subsection 245(1) of the Act because, after the transaction, Canada Trustco The background to these rules is discussed in the judgment of Miller J. in the Tax Court, Canada Trustco, supra note 80 (T.C.C.) at 595. 4 2 1 The 1989 amendments, as well as their rationale, are also discussed by Miller J., ibid, at 595 and 602-03. 154 422 was entitled to a tax deferral to which it had not previously been entitled. In addition, all levels of court agreed that the transaction was an avoidance transaction under subsection 245(3). The transaction was clearly undertaken for tax purposes, since Canada Trustco had sought out a transaction that would obtain capital cost allowance for the purpose of sheltering other taxable income. The circular financing sale-leaseback that Canada Trustco engaged in is a classic avoidance transaction. Finally, with respect to subsection 245(4), the Crown argued that the transaction resulted in a misuse or abuse because Canada Trustco acquired legal ownership of assets entitling it to millions in capital cost allowance without incurring any economic cost or sacrifice to purchase the assets. The Crown argued that the object and spirit of the capital cost allowance rules generally is to recognize the cost of money spent to acquire capital assets for the purpose of earning income. The Crown relied on the decision in Water's Edge Village Estates (Phase II) Ltd. v. i ? . , 4 2 3 another G A A R case considering the capital cost allowance provisions. In Water's Edge, the Federal Court of Appeal had characterized the object and spirit of the capital cost allowance provisions as "to recognize money spent to acquire qualifying assets." 4 2 4 In effect, the Crown argued that the object and spirit of the capital cost allowance provisions presumes that the cost of an asset is a real economic cost, and that the circular financing sale-leaseback resulted in a misuse or abuse because it manufactured an illusory cost that was not actually incurred in economic terms. 4 2 5 Refer to Chapter 5, note 344, for comments regarding whether the tax benefit resulting from this transaction should be characterized as a tax deferral or a tax reduction. 4 2 3 2002 D.T.C. 7172 (F.C.A.) [Water's Edge]. 4 2 4 Ibid, at 7179. 4 2 5 The Crown's argument was somewhat impaired by the fact that the Crown advanced a different figure for Canada Trustco's actual economic cost of the trailers at each level of appeal. 155 None of the courts that heard Canada Trustco agreed that the G A A R applied. The Supreme Court of Canada concluded that the transaction did not frustrate Parliament's intent. The Court rejected the notion that the capital cost allowance provisions presume that a real economic cost to acquire assets. Applying its approach to object and spirit, as explained in Chapter 5, the Court stated that economic substance is only relevant i f the wording of a specific provision suggests that the object and spirit of that provision intends to take economic substance into account. 4 2 6 The Court observed that nothing in the statutory definition of "cost" suggests that economic substance is a relevant consideration. The Court held that "'cost' in the context of [the capital cost allowance] provisions is a well-understood legal concept [that] has been carefully defined by the Act and the jurisprudence." 4 2 7 The Court did not agree that subsection 245(4) imports a general economic substance requirement into the Act as a whole. Therefore, the Court reasoned, the object and spirit of the provisions 498 is only to recognize Canada Trustco's legal cost. Since the sale, lease and loan transactions were all valid and legally binding, the legal substance and reality of the transaction was that Canada Trustco paid $120 mill ion plus fees for the trailers. The Court also agreed with the Tax Court that the circular financing sale-leaseback using borrowed funds was not "so dissimilar" from an ordinary sale-leaseback (which Parliament permits through the specified leasing property rules and the exemption for listed property) "as to take it outside the object, spirit or purpose of the relevant [capital cost allowance] provisions." 4 2 9 In an ordinary sale-leaseback, the purchaser/lessor would take on some risk in acquiring the assets, and would have an economic cost. According to the Tax 4 2 6 Refer to the discussion of economic substance under the GAAR in Chapter 5 , under the heading "An Exception for Misuse, Abuse and Frustration." 427 Canada Trustco, supra note 80 at 5535, para. 75. 428 Ibid, at 5535, paras. 74-75. 429 Ibid, at 5536, para. 78. 156 Court Judge, however, all that distinguished the Canada Trustco transaction from ordinary sale-leaseback financing was that Canada Trustco effectively minimized its economic r i sk . 4 3 0 The complete absence of risk resulting from the circular flow of funds did not create a fundamentally different transaction from an ordinary sale-leaseback. i i . Normative Tax Consequences and Applying the G A A R Canada Trustco is the type of tax avoidance that the G A A R was intended to prevent. The courts correctly concluded that the transaction resulted in a tax benefit and was a tax-driven avoidance transaction. However, the courts wrongly concluded that the transaction was not abusive and did not frustrate Parliament's intent. In order to explain why the G A A R should have applied to the transaction, it is first necessary to identify why the transaction is abusive and what the normative tax consequences of the transaction would be - in other words, how Parliament intended or would have intended to treat the transaction. In Chapter 3,1 argued that tax legislation is merely a policy instrument through which Parliament imposes tax outcomes in pursuit of broader fiscal and economic policies. I also suggested that, unless Parliament indicates otherwise, tax consequences are intended to be based on taxpayers' economic circumstances. In my view, when the nature of tax legislation is properly understood, tax policy considerations make an economic cost requirement in a situation such as Canada Trustco apparent. From a tax policy perspective, capital cost allowance is intended to recognize the cost of assets used to earn income, and to properly Ibid, at 5536, para. 78. match that cost with revenues over an asset's useful lifetime. Therefore, tax policy considerations suggest that, in order to recognize an acquisition of assets for capital cost allowance purposes, a taxpayer ought to have incurred an economic cost in the hope of generating an economic benefit against which that cost can be matched. A transaction that does not involve any economic cost, either in the form of capital laid out or risk assumed, is not the type of transaction that Parliament intended to recognize for capital cost allowance purposes. In my view, therefore, the normative tax treatment of the Canada Trustco transaction would be not to recognize the transaction as an acquisition of capital assets for capital cost allowance purposes because it is not the type of acquisition that Parliament intended to recognize. The contrary argument, that Parliament did not intend to require any economic cost and that legal cost alone is determinate, becomes absurd if taken to the extreme. Had Canada Trustco needed to shelter more than the $50 mil l ion that it did, it could simply have structured a transaction using different amounts than those used. In fact, using borrowed funds immediately recycled to the lender, Canada Trustco likely could have manufactured a similar transaction resulting in any amount of capital cost allowance that it desired, as long as assets could be found in the marketplace. Since the purchase would be funded with borrowed funds immediately recycled to the lender, Canada Trustco could have purchased two or even three times the value of assets at no additional cost to itself. It is unreasonable to suggest that Parliament intended to permit taxpayers to benefit from this type of artificial 4 3 1 It is also arguable that, to the extent that capital cost allowance rates exceed the actual rate of depreciation, Parliament offers businesses a form of subsidy to encourage capital investment in the pursuit of income. However, I do not believe that even this purpose is satisfied by Canada Trustco, since the trailers were not acquired in the pursuit of income but in pursuit of tax deductions. The transaction was structured so that it automatically offset any leasing income generated by the trailers with notional interest expenses paid to the bank. 158 transaction that could be manipulated to engineer whatever amount of tax benefits taxpayers require. In light of this factor, an argument that Parliament intended the capital cost allowance provisions to focus only on legal cost, to the exclusion of economic considerations, is absurd. In terms of the theoretical formulation of tax avoidance outlined in Chapter 3, Canada Trustco is abusive tax avoidance because it involves a tax-driven transaction that engineered a tax result based on legal relationships that did not reflect the underlying economic outcome that Parliament would have intended to exist in order to entitle Canada Trustco to capital cost allowance. The transaction was decidedly not an example of legitimate tax mitigation. Canada Trustco incurred neither the cost, nor the economic sacrifice, to acquire the trailers that Parliament would have intended. Nor, even though capital cost allowance does sometimes function as a tax incentive, 4 3 2 does the purchase of trailers at no economic cost for the purpose of deducting capital cost allowance qualify as legitimate tax mitigation. In reviewing the object and spirit of the capital cost allowance provisions, the Court wrongly concluded that those provisions do not require any consideration of economic cost. The Court focused only on the statutory definition of "cost" and the absence of any words explicitly invoking economic cost. This conclusion is based on the assumption that, unless tax legislation explicitly provides otherwise, only legal relationships are relevant for the purpose of determining tax consequences. Evolving from the Duke of Westminster principles of statutory interpretation, this assumption is expressed most clearly in Shell, where the Court 4 3 2 Such as when Parliament accelerates the capital cost allowance rate specifically to encourage changes in behaviour, such as by granting a 100% write-off for Y2K-compliant computers or by accelerating rates to encourage investment in green technology. As explained in the previous note, these features of the capital cost allowance provisions are nonetheless aimed at real, actual economic transactions, not merely at artificial tax-driven ones. 159 observed that, "absent a provision to the contrary," economic realities are not relevant. The G A A R is such a provision to the contrary because it denies tax benefits where transactions result in a misuse or in an abuse of the Act. The reference to the "result" of transactions is the signal that the economic circumstances underlying transactions are relevant in determining whether they are abusive. In Canada Trustco, the Court framed the subsection 245(4) analysis as a question of whether giving the impugned transaction effect for tax purposes would frustrate the intent of Parliament. It would be impossible to properly answer that question without recalling that Parliament's overriding intention is to impose tax consequences based on underlying economic outcomes. The Court also wrongly concluded that the object and spirit of the capital cost allowance provisions permit the Canada Trustco transaction because it was not significantly different from an ordinary sale-leaseback not using circular financing. The Court failed to appreciate that the circular flow of funds distinguishes Canada Trustco from the types of lease financing that Parliament intends to permit. A "normal" sale-leaseback would involve an expenditure of funds by the purchaser-lessor. Those funds would be at risk until the lease was paid off. They would not be placed in a secure investment pledged back to the purchaser as security, nor would they be guaranteed a rate of return. Even if the purchaser-lessor financed the purchase with borrowed funds, there would still be a real liability to the lender. The purchaser-lessor would also assume the risk of default by the vendor-lessee, or that the assets would be destroyed and they would be left to repay the loan with nothing to show for it. The borrowed funds would not have been recycled to the lender, to sit in an account held by the lender's controlled subsidiary. Thus, the circular flow of funds distinguishes the Canada Trustco transaction from a "normal" sale-leaseback in significant ways and 433 Supra note 99 at 5676-77. 160 demonstrates how economically unreal it is. A t the end of the transaction, the parties, assets, and money had all returned to where they had started; and it was pre-ordained that the parties who had advanced funds would be repaid. The only activity that actually occurred was a paper shuffle that altered legal relationships to transfer the legal entitlement to capital cost allowance. It is possible to apply the G A A R in a manner that identifies abusive tax avoidance and that ensures the normative tax consequences that Parliament would have intended. To do so merely requires that a court applying the G A A R acknowledge that a consideration of economic substance is an essential component of the abuse analysis. Such an approach has even been followed before. Water's Edge,434 which also concerned a claim for capital cost allowance, is one example. In Water's Edge, a US partnership owned a mainframe computer originally purchased for U S $3.7 mill ion (about Cdn. $4.5 million), but which had been fully depreciated for U S tax purposes and which had a market value of only U S $7,000. The taxpayer and other investors acquired 93.5% of the U S partnership for $320,000 and caused it to become a 50% partner in a new Canadian partnership. The US partnership contributed the computer to the new partnership and was credited $50,000. The U S partnership then claimed a terminal loss of just under $4.5 mill ion on the basis that a depreciable asset with a cost of $4.5 mill ion had been disposed of for proceeds of $50,000. The taxpayer, having paid a share of the $320,000 spent to acquire 93.5% of the U S partnership, then claimed its proportionate share of the terminal loss under the provisions flowing partnership income and losses through to its members. The Federal Court of Appeal concluded that the G A A R applied to this transaction. The Court found the transaction abusive because it exploited a loophole whereby the 434 Supra note 423. 161 computer's historical cost was its cost for Canadian tax purposes, without any adjustment to recognize either its market value or the fact that it had previously been depreciated in another jurisdiction. The Court concluded that the object and spirit of the capital cost allowance provisions is to "provide for the recognition of money spent to acquire qualifying assets to the extent that they are consumed in the income-earning process under the A c f . " 4 3 5 Although the Court did not specifically refer to economic cost, the Court clearly considered the economic result of the transaction and even referred to the "actual cost" of a capital asset. 4 3 6 In doing so, the Court implicitly recognized that the actual economic cost of the computer to the taxpayer was no more than its share of the $320,000 paid to acquire 93.5% of the U S partnership and was nowhere near the cost amount claimed. As in Canada Trustco, the abusive tax avoidance in Water's Edge resulted from the taxpayer legally obtaining capital cost allowance (in this case, a terminal loss) deductions without incurring a corresponding real economic cost to acquire the asset. Water's Edge indicates that the G A A R can be applied to tax avoidance transactions in a manner that preserves normative tax consequences. Even though the Court did not explicitly refer to economic cost, the underlying economic circumstances cast their spectre over the entire case. The Court could not accept that Parliament intended to give effect to a transaction resulting in a legal loss of $4.5 mill ion at an economic cost of only $320,000. From a tax policy perspective, all that the Court could have done better in Water's Edge is to be more explicit in acknowledging that the economic circumstances were an important factor in evaluating the result of the transaction as required by subsection 245(4). Water's Edge, supra note 423 at 7179, para. 44. Ibid, at 7179, para. 44. 162 Finally, the fact that sophisticated taxpayers like Canada Trustco can pay advisors to manufacture tax benefits on demand illustrates the concerns about equity as between taxpayers highlighted in Bronfman Trust, but later discarded by the Supreme Court of Canada in Shell.431 Since equity is an essential feature of a good tax system, tax policy considerations require the G A A R to be interpreted in a manner that encourages equity by disallowing artificial, contrived tax benefits that less sophisticated taxpayers with fewer resources do not have an equal opportunity to obtain. Otherwise, the tax system wi l l become a free-for-all in which those with the resources and the opportunity to engage in sophisticated tax avoidance schemes wi l l bear less than their appropriate share of the tax burden, thus undermining equity and tax fairness. B. Surplus Stripping: McNichol and Evans i . McNichol: The Facts and Judgment McNichof3S and Evans v. The Queen439 illustrate a tax phenomenon known as "surplus stripping," in which shareholders realize their interest in a corporation by selling their shares to receive capital gains rather than by simply withdrawing the corporate surplus as dividends. Taxpayers engage in surplus stripping because the effective tax on capital gains is lower than that on dividends. McNichol involved four equal shareholders in Bee Holding Corporation Limited ("Bee") that owned a building from which it earned rental 4 3 7 Refer to Chapter 4, under the heading "Substance Over Form in the UK and Canada," and to Chapter 5 under "Statutory Interpretation and Tax Avoidance in Canada" for a discussion of Bronfman Trust and the equity concerns raised by tax avoidance. 438 Supra note 90. 4 3 9 2005 D.T.C. 1762 (T.C.C.). 163 income. Bee sold the building for $600,000 and distributed the non-taxable portion of its capital gain to the shareholders as a capital d ividend. 4 4 0 After the sale, there was no prospect that Bee would continue carrying on business. Its assets consisted of approximately $319,000 in cash and $34,000 of refundable dividend tax on hand. 4 4 1 The four shareholders then decided to end their association in the corporation. Rather than wind up Bee and declare a dividend to each shareholder, the shareholders sold their shares of Bee to Beformac Holdings Limited ("Beformac"), a new corporation formed by one of their clients, for $300,000. Beformac had no funds and purchased the shares with $300,000 borrowed on the security of Bee's cash. After the share purchase, Bee and Beformac were amalgamated and Beformac repaid its bank debt with Bee's cash. The prior shareholders of Bee reported capital gains of $75,000 each 4 4 2 and claimed their capital gains exemptions. 4 4 3 Bee's shareholders could have simply wound up Bee and split the corporate surplus by equally by declaring a dividend to each shareholder. However, the shareholders instead elected to arrange a share sale to realize their interest in the corporate surplus, because the resulting capital gains were more tax-efficient than dividends due to the lower inclusion rate and the lifetime capital gains exemption. Therefore, they located a purchaser by offering a share of the tax savings that would result from selling shares instead of dividends. 4 4 4 There appeared to be no business reason for Beformac to purchase the shares of Bee, and in fact Beformac could only do so with a transitory loan that it later repaid with Bee's cash. The 4 4 0 A capital dividend is a corporation's distribution of the non-taxable portion of its capital gain to the shareholders as a dividend. Under subsection 83(2) of the Act, a capital dividend is tax-free to the shareholder receiving it. 4 4 1 Section 129 of the Act permits a private corporation that pays a taxable dividend to obtain a partial refund of the income taxes that it paid on its investment income. Under subsection 129(3) of the Act, refundable dividend tax on hand is a notional account that represents the current value of anticipated refunds. 4 4 2 Each shareholder's adjusted cost base in their Bee shares was only $125. 4 4 3 Refer to note 330 regarding the "lifetime capital gains exemption" available under section 110.6 of the Act. Two of the partners thereby received their $75,000 share of Bee's surplus entirely tax-free. 4 4 4 It appears that this is why the shares were sold for less than the actual cash in Bee. 164 only way in which the purchase benefited Beformac was the surplus cash remaining after repaying the bank loan. The Crown invoked the G A A R , arguing that the share sale was abusive tax avoidance and that the shareholders should instead be taxed as i f they had received the corporate surplus through dividends from Bee . 4 4 5 The Crown identified the tax benefit as the reduction in tax payable by the shareholders by virtue of their receiving exempt capital gains instead of dividends. The Crown argued that the transaction was an avoidance transaction because it was undertaken primarily for the purpose of reducing the tax payable on the extraction of Bee's economic surplus. Finally, the Crown argued that the transaction amounted to a misuse and abuse because the Act does not countenance surplus stripping. Rather, the Crown argued, the Act intends distributions of corporate surplus to be taxed as income to the shareholders. The Tax Court agreed that the G A A R applied. The Court agreed that identifying a tax benefit in this case required a comparator transaction, and that the appropriate comparator transaction was one where the tax consequences would have been those that resulted from a dividend. The Court held that it was not reasonable to assume that the shareholders would have done nothing had they not sold their shares. Since their ultimate objective was to realize the economic value of Bee's surplus, it was reasonable to conclude that they would have declared a dividend had they not sold their shares. 4 4 6 Therefore, because the transaction 4 4 5 The Crown made two other arguments based on specific rules relating to the taxation of dividends. The first was that the payment by Beformac to the shareholders was a distribution of Bee's funds on the winding-up of its business and therefore a deemed dividend under subsection 84(2) of the Act. The second was that the payment was a deemed dividend under section 84.1 of the Act because the shareholders did not deal at arm's length with Beformac at the time of the sale. Neither argument succeeded. The Tax Court held that the sale of shares to Beformac could not be regarded as a distribution of Bee's funds because, legally, it was a purchase by Beformac using its own funds, albeit borrowed funds. The Tax Court held that section 84.1 did not apply because the parties were dealing at arm's length and bargaining in pursuit of their own economic interests. 4 4 6 McNichol, supra note 90 at 119. 165 resulted in lower tax payable than a dividend would have, the transaction resulted in a tax benefit under subsection 245(1) of the Act. 4 4 7 The Court also agreed that the transaction was an avoidance transaction under subsection 245(3) of the Act. The Court explained this conclusion on the basis that the share sale transaction was not selected for bona fide reasons but to give rise to a tax-exempt capital ga in . 4 4 8 Finally, the Court concluded that the transaction resulted in a misuse or abuse because the scheme of the Act contemplates that "distributions of corporate property to shareholders are to be treated as income in the hands of the shareholders." 4 4 9 The Court identified the "scheme of the Act" by considering sections such as paragraph 12(l)(j), section 15, and section 84, which deal with various ways in which shareholders might receive corporate property and treat those receipts as income. 4 5 0 Upon reviewing this legislative scheme, the Court concluded that the form of such distributions is generally irrelevant to the intention to treat them as dividends. 4 5 1 The Court observed that the transaction "[took] 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." 4 5 2 The Court appears to have been saying that the result of the transaction was a distribution of corporate surplus even though the legal form was a sale of shares. The Court therefore concluded that the shareholders' tax consequences should be redetermined as if they had each received a $75,000 dividend. 441 Ibid, at 119. 4 4 8 In other words, the primary purpose of the transaction was to obtain a tax benefit - i.e., to bring about a lower tax liability than a dividend would have done. Ibid, at 119. 4 4 9 Ibid, at 121-22. 4 5 0 Paragraph 12(l)(j) requires that actual dividends be included in income. Section 15 requires that shareholder appropriations of corporate property, or other benefits bestowed on shareholders, as opposed to as dividends or salary, are included in income. Section 84 deems dividends to have been paid in a variety of situations. Finally, the Court observed that subsection 83(2) identifies specific circumstances in which a distribution of corporate property to shareholders is excluded from income, and characterized that provision as the exception to a general rule. 451 McNichol, supra note 90 at 120. 4 5 2 Ibid, at 122. 166 i i . Evans: The Facts and Judgment McNichol was decided in 1997, eight years before Canada Trustco. Evans, on the other hand, was decided just a month after Canada Trustco, and Evans rejected the McNichol reasoning. In Evans, the taxpayer was a dentist practising through a professional corporation, 117679 Alberta Ltd. ("117679")- In addition, the corporation also operated a dental hygiene business and rental properties. Through a series of tax-free rollovers, Evans rolled the dental practice and rental business into two new corporations, leaving only the hygiene business in 117679. 117679 then declared a $487,000 stock dividend to Evans, which stock he sold for $487,000 to a limited partnership consisting of his wife as a 1% partner and three of his children as 33% partners each. The partnership did not pay for the shares, but instead issued Evans a promissory note. Evans declared a capital gain of $486,900 on the sale of the shares, and sheltered it under his lifetime capital gains exemption pursuant to section 110.6 of the Act. Over three years, 117679 paid the partnership more than $267,000 through dividends and redemptions of the shares purchased from Evans. The partnership paid those same amounts to Evans as principal and interest on the promissory note. No profits were kept by the partners and all of the funds ended up in Evans' hands. 4 5 3 A s partners in the partnership, Evans' family members included the value of the dividends and deemed dividends in their income in those years. However, they were not taxable in those years, and the only receipts that were taxable in Evans' hands were the amounts characterized as interest on the promissory note from the partnership. Amounts that Evans received as repayments of principal were not taxable. 4 5 3 $158,113 was characterized as repayments of principal owed to Evans. 167 The Crown relied on the G A A R , arguing that the transaction was a surplus strip whereby amounts that would have been dividends included in Evans' income were directly to a non-arm's length partnership whose members were not taxable, and then were received by Evans in the form of non-taxable principal repayments. The Crown argued that the tax benefit was the reduction in tax to Evans resulting from his receiving corporate surplus as repayments of principal rather than as dividends, and that the primary purpose for the stock dividend and the sale of shares to the partnership in exchange for a promissory note was to obtain that tax benefit. The Crown argued that the series of transactions resulted in a misuse or abuse for the same reasons that the sale of shares in McNichol was abusive. In response, Evans contended that the transactions were undertaken for the bona fide purpose of estate planning and transferring wealth to his family members. Unlike in McNichol, the Tax Court did not agree that the G A A R applied to these transactions. In identifying a tax benefit, the Court applied the same analysis as in McNichol. The tax benefit was a reduction or avoidance of the tax that Evans would have paid had he received the $267,000 in dividends directly, as compared to the tax that he actually paid on the amounts received as interest on the promissory note. 4 5 4 However, the Court differed from McNichol on the question of the primary purpose of the transactions. The Court rejected Evans' argument that the purpose of the transactions was to transfer wealth to his family members since all of the money ultimately flowed to Evans himself. The Court characterized the goal of putting corporate funds in Evans' hands as an "economic objective." The Court then described the transaction as "a non-tax purpose of getting corporate funds in Dr. Evans' hands and a method of doing so by means of a series of transactions whose primary purpose Evans, supra note 439 at 1766, para. 17. 168 - arguably only purpose - was to enable him to do so at a significantly reduced tax cost." 4 5 5 Based on the conclusions that the primary purpose of the series of transactions - putting the corporate funds into Evans' hands - was a valid economic objection, and that the transactions were only a method of doing so at a reduced tax cost, the Court concluded that the transactions were not avoidance transactions. 4 5 6 In so deciding, the Court relied on the Explanatory Notes, which state that subsection 245(3) does not permit the "recharacterization" of transactions and "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." 4 5 7 The Court continued and concluded that, even i f the transactions were avoidance transactions, they did not result in a misuse or abuse of the Act. After observing that Evans legally complied with each of the specific rules relied upon in structuring his transactions, 4 5 8 the Court rejected what it characterized as a suggestion that "there is some overarching principle of Canadian tax law that requires that corporate distributions to shareholders must be taxed as dividends." The Court held that Canada Trustco explicitly prohibits ignoring specific sections of the Act.m Finally, the Tax Court distinguished this case from McNichol 455 Ibid, at 1766, para. 20. 456 Ibid, at 1767, para. 22. 457 Supra note 364 at 2265. 4 5 8 The Court listed the following provisions that Evans utilized in his tax plan, and with which he literally complied: section 85 (the rollover of assets to a corporation), section 110.6 (the lifetime capital gains exemption), subsection 84(3) (deemed dividends on the redemption of shares), section 112 (tax-free intercorporate dividends), subsection 52(3) (stock dividends), and section 74.5 (an exception from the attribution rules where fair market value consideration is paid). 4 5 9 In a brief comment at the end of the decision, the Court also commented that the transactions did not lack economic substance, not because they had a business purpose, but because a genuine change in legal and economic relations occurred. Evas, supra note 439 at 1770-71, para. 35. However, the Court did not appear to seriously consider the meaning of economic substance. 169 on the basis that, unlike the Bee corporation in McNichol, 117679 continued carrying on business and actually paid out dividends after the transactions. 4 6 0 i i i . Normative Tax Consequences and Applying the G A A R Both McNichol and Evans present similar scenarios in which shareholders who would have received distributions of corporate profits as dividends entered into transactions that resulted in them receiving corporate surplus in a different legal form that attracted lower tax. In McNichol, the sale of shares to Beformac, using borrowed money effectively funded by Bee's surplus, resulted in a capital gain instead of a dividend. In Evans, the corporate surplus flowed through a partnership with low tax-rate members and was then paid to the shareholder as repayments of principal under a promissory note created to enable the funds to move to the shareholder tax-free. A l l of the transactions in McNichol and Evans were valid and legally effective. Their legal substance was that they resulted in capital gains to the McNichol shareholders and debt repayments to Evans. A l l of the transactions also complied with the provisions in the Act dealing with capital gains, stock dividends, partnership taxation, and debt repayments. However, neither case presents a transaction that Parliament anticipated in enacting the legislative provisions dealing with dividends and capital gains. Therefore, when considering whether the G A A R should apply to these transactions, the questions to be asked are what tax consequences would Parliament have intended to result from these transactions and whether the transactions actually undertaken frustrated that intent. In my view, the McNichol decision appropriately identifies the normative tax consequences, based on its review of the scheme of the Act as it relates to distributions of 460 Ibid, at 1770, para. 34. 170 corporate property to shareholders. The fact that paragraph 12(l)(j) requires formal dividends to be included in a shareholder's income, subsection 84(2) deems dividends to have been paid in a variety of situations that do not include the formal declaration of a dividend, and subsection 15 requires other shareholder receipts of property from a corporation to be included in income as shareholder benefits, strongly suggests that Parliament has attempted to create a complete code for the taxation of corporate property received by shareholders. 4 6 1 Only the provisions permitting a return of capital and the payment of capital dividends provide exceptions to Parliament's general intent of taxing corporate distributions as dividends. Contrary to the Tax Court's comments in Evans, therefore, the scheme of the Act in fact does disclose an "overarching principle of Canadian tax law that requires that corporate distributions to shareholders must be taxed as dividends," whatever their form and subject to specific exceptions for the return of capital and the flowing-through of the non-taxable portion of corporate capital gains as capital dividends. In McNichol, the share sale transaction was not undertaken for any business purpose. It was a relatively transparent attempt to wind up Bee without legally winding up and declaring a dividend. However, the transaction nonetheless resulted in a distribution of the surplus of Bee, which ceased to exist when it amalgamated with Beformac, a transitory corporation that carried on no business and served no purpose other than facilitating a surplus strip. Had the Bee shareholders sold their shares to a new purchaser intent on owning Bee and carrying on its business, Parliament would not have intended the resulting capital gain to be treated as a dividend. That tax treatment would not be intended because such a sale would 4 6 1 This analysis could be expanded to include the fact that payments to shareholder for services rendered, whether as employment or business income, would also be income to the shareholders. Furthermore, the presence of section 84.1, which creates deemed dividends on a sale of shares to a non-arm's length corporation, could be added to the list of provisions considered by the Tax Court to form the "code" for taxing distributions of corporate surplus as income. 171 not have resulted in a distribution of the surplus of Bee, but in a disposition of the business to a new owner. The context in which the McNichol share sale occurred is therefore an important factor in determining the normative tax consequences. The choice to sell to a transitory purchaser instead of winding up and declaring a dividend was a choice between competing legal structures for achieving the economic outcome of distributing corporate surplus. Therefore, in my view, the normative tax treatment that Parliament would have intended to result in McNichol is to treat the gains received by the shareholders as dividends. 4 6 2 Similarly, in Evans, the series of transactions that resulted in corporate surplus being received by the shareholder as repayments of debts from the partnership rather than as dividends from 117679 served no purpose other than extracting the surplus of 117679 in the most tax-efficient manner. In fact, the Court itself acknowledged that the transactions had no other purpose 4 6 3 The Court also acknowledged that the members of the partnership that legally received the dividends never benefited from that income. In effect, the partnership functioned merely as a conduit so that 117679's corporate surplus would be legally received by low-rate taxpayers before flowing through to Evans. In my view, the normative tax treatment of the Evans transactions is no different from that of the McNichol transactions. The fact that 117679 continued its business and paid dividends does not distinguish this case from McNichol. The legislative scheme identified in McNichol is clear and it makes no difference that the Evans transactions resulted in corporate surplus being received in a lower-taxed form on an ongoing basis as opposed to in the context of a winding-up. Parliament's 4 6 2 In addition, it is evident that the transactions did not result in a distribution of capital, which would not be taxable to the shareholders, since Bee had already paid the capital dividend resulting from its gain on the sale of its building even before the share sale occurred. 4 6 3 Evans, supra note 439 at 1766, para. 20. 172 intention remains the same - to treat distributions of corporate surplus, except returns of capital and distributions of the exempt portion of corporate capital gains, as income to the shareholders. This intention holds true whether the distribution is an annual dividend out of retained earnings or a distribution of all corporate property on winding-up. Therefore, the normative tax treatment of Evans would be to tax him as i f he had received a dividend despite the existence of the partnership and promissory note. The fact that the Tax Court reached different conclusions in McNichol and Evans is disturbing, given that the result of each transaction is substantially the same - amounts that would have been received as dividends by the shareholder were received in another legal form that attracted less tax. I have suggested above that there is no principled reason for Evans' transactions to attract different tax consequences than the tax outcome that Parliament appears to have intended to apply to such an economic result. In my view, this difference resulted because, although the Tax Court appropriately applied the G A A R in McNichol, it made two key errors in Evans. In both cases, the Court applied the appropriate test for identifying a tax benefit. However, in Evans the Court first misapplied the test for an avoidance transaction by confusing the transactions' purposes with the taxpayer's objective or motivation, and then failed to appreciate the object and spirit of the provisions of the Act dealing with corporate surplus. Those provisions, as explained in McNichol, contemplate taxing such distributions as dividends. In both cases, the Court acknowledged that the transactions were undertaken in pursuit of the goal of removing corporate surplus and putting it into the shareholders' hands. Therefore, the appropriate way to identify a tax benefit under subsection 245(1) of the Act is by considering what the shareholders would reasonably have done to extract the surplus had they not undertaken the transactions in issue, and to use those 173 alternative transactions as the appropriate comparators. Those alternatives are the declaration of a dividend in McNichol and the payment of dividends directly to Evans in Evans. Since the transactions actually undertaken resulted in the shareholders receiving corporate surplus at a lower tax rate than the reasonable comparator, the transactions resulted in tax benefits. With respect to the purpose test in subsection 245(3), the Court properly applied the test in McNichol but misapplied it in Evans. In concluding that the primary purpose of Evans' transactions was to put corporate surplus into his hands, the Court focused on Evans' subjective motivation or economic objective as opposed to the objective purpose of the transactions. Subsection 245(3) requires a determination of what the purpose of the transactions can reasonably be considered to be. In Evans, even the Court, at one point, recognized that the only purpose of the transactions - stock dividend, partnership, promissory note, and repayments of principal - was to flow the corporate surplus to Evans at a lower tax cost than regular dividends. 4 6 4 Objectively speaking, the purpose of the transactions, in the sense of the result that they were designed to bring about, was to have corporate surplus received first by low-rate taxpayers and then to transfer it tax-free to Evans as a repayment of principal. The only reasonable purpose of the transactions, therefore, was to reduce the tax cost of achieving Evans' ultimate economic objective 4 6 5 In deciding as it did, the Tax Court made the same error as made in Canadian Pacific, the weak currency loan case. In that case, the Federal Court of Appeal held that the 4 6 4 Ibid, at 1766, para. 20. 4 6 5 In addition, the decision is internally inconsistent. At paragraph 19, the Court rejected Evans' argument that the purpose of the transactions was to transfer wealth to his children. At paragraph 20, the Court held that the "non-tax purpose" of the transactions was to get corporate funds into Dr. Evans' hands and that the series of transactions utilized to do so had a primary purpose, or even arguably a sole purpose, of doing so at a significantly reduced tax cost. In virtually the same breath, therefore, the Court concludes that the transactions have a non-tax purpose and that their primary, if not sole, purpose is tax reduction. The only way to avoid the internal contradiction in this passage is to characterize the receipt of corporate funds as the taxpayer's "non-tax objective" as opposed to as the purpose of the transactions themselves. Ibid, at 1766, paras. 19-20. 174 intermediate borrowing of Australian dollars, followed by the conversion of those dollars into Canadian dollars, was not an avoidance transaction. The Court concluded that the primary purpose of all of the borrowing steps was to obtain capital for business use. 4 6 6 The Court adopted the proposition that "no transaction forming part of the overall series can be viewed as having been arranged for a purpose which differs from the overall purpose of the series." 4 6 7 Unfortunately, the decision in Canadian Pacific is plainly wrong. It is wrong because the definition of avoidance transaction in subsection 245(3) clearly distinguishes between a series of multiple transactions and the individual transactions that comprise the series 4 6 8 Each individual transaction step in a series has its own objective purpose, and it is that purpose, not the purpose or objective of the series as a whole, that must be tested under subsection 245(3). 4 6 9 A s explained in Chapter 5, defining the purpose of each step in the series in terms of the purpose of the overall series confuses the objective purpose of each transaction - the result that the transaction is designed to bring about - with the taxpayer's ultimate subjective objective in pursuing the series itself. 4 7 0 Furthermore, it is clear from the structure of subsection 245(3) itself and from the Explanatory Notes that the G A A R is intended to deal with step transaction in which tax-reducing steps are inserted into series of transactions that themselves achieve non-tax goals. To read subsection 245(3) without making this distinction is to read the step transaction aspect of the G A A R out of existence. 4 7 1 Canadian Pacific, supra note 99 at 6749, para. 27 467 Ibid, at 6749, para. 27. 4 6 8 Arnold, "Long, Slow Steady Demise", supra note 58 at 495. 4 6 9 Ibid, at 495. 4 7 0 Refer to Chapter 5 under the heading "Objective Purposes as Compared to Subjective Motives," and to note 358 above. 4 7 1 Arnold, ibid, at 495, argues that such a reading would make the GAAR "largely meaningless" because it would apply only to "solitary" transactions and series of transaction where the entire series is undertaken in pursuit of a tax benefit. 175 In addition to failing to properly identify the objective purpose of Evans' transactions, the Court failed to appreciate the object and spirit of the statutory provisions that intend to treat distributions of corporate surplus, regardless of their form, as dividends. The Court rejected the McNichol decision on the basis of a factual distinction that is logically immaterial. The only real argument in support of the Court's decision is that 117679 did legally distribute its surplus as dividends before those amounts were paid to Evans, and that they were therefore taxed in the hands of the recipients in accordance with the scheme of the Act. However, this argument ignores the fact that the result of the transactions is that all corporate surplus went to Evans. No funds remained in the partnership or were distributed to the partners; all were paid to Evans to repay the promissory note created specifically for the purpose of characterizing payments to him as repayments of principal. As I argued in discussing Canada Trustco above, the phrase "results in a misuse or abuse" in subsection 245(4), if it is to have any meaning, must be interpreted by reference to the underlying economic result or outcome of a transaction. Subsection 245(4) must be interpreted this way 479 because, unless the words or the scheme of the legislation indicates otherwise, Parliament generally intends to impose tax consequences on the basis of taxpayers' underlying economic outcomes and circumstances. Otherwise, sophisticated taxpayers like Canada Trustco and Evans wi l l be free to manufacture legal relationships on demand in order to obtain favourable tax consequences, with the associated detrimental effects on equity and fairness between 473 taxpayers. 4 7 2 Such as in the case of tax incentives that should be available to transactions without a non-tax purpose and that do not result in the intended economic outcomes. 4 7 3 McDonnell suggests that, by using a partnership instead of a corporation, Evans sidestepped the potential application of section 84.1, which would have deemed him to receive dividends had a corporation controlled by his family members acquired the shares and flowed the corporate surplus to him. The selection of a partnership instead of a corporation might amount to an abuse of a legislative scheme intended to prevent taxpayers from In terms of the theoretical formulation of tax avoidance proposed in Chapter 3, the McNichol and Evans transactions are abusive tax avoidance because the taxpayers succeeded in re-casting the receipt of corporate surplus in a legal form that attracted less tax without altering the economic result of their transactions. As a result, the tax consequences that the legislation would have imposed on their transactions had the G A A R not applied would not have been those that, from a tax policy perspective, Parliament intended to apply. Furthermore, the transactions were abusive because they were not undertaken for any purpose other than tax-reduction, and the scheme of the legislation clearly indicates that Parliament did not intend that result. McNichol illustrates that the G A A R can be applied in a manner that preserves the normative tax outcomes for a surplus stripping transaction. , However, such an application requires that courts appreciate the tax-reduction purposes of the transactions used to achieve a non-tax objective (extracting corporate surplus), and that courts distinguish between those concepts. It also requires that courts appreciate the intention of Parliament to tax distributions of corporate surplus as dividends, as is apparent from the scheme of the Act. As McNichol points out, Parliament intends to do so regardless of the legal form in which that surplus is received. Unfortunately, the Tax Court appears to have treated the Canada Trustco decision as an invitation to abandon the ready-made expression of the object and spirit of the Act, and of Parliament's intent, found in McNichol and to allow abusive tax avoidance. doing what Evans did. Thomas E. McDonnell, "The Evans Case: GAAR in a Post-Canada-Trustco World" (2006) 6:1 Tax for the Owner Manager 9 at 10. 177 C. Conclusions This Chapter has considered two actual tax avoidance scenarios - a circular financing sale-leaseback and surplus stripping - which have been challenged in Canadian courts, and to which the tax authorities have not been particularly successful in applying the G A A R . Applying the tax-policy based theory of tax avoidance articulated in Chapter 3, both scenarios indeed are abusive because they result in the legislation applying tax consequences (based on the legal relationships created in the transactions) that are not the tax consequences that, from a tax policy perspective, Parliament would have intended to apply. Furthermore, both scenarios involve artificial transactions arranged for the purpose of securing tax reductions by creating legal forms that diverge from the economic outcomes of the transactions. Thus, the approach articulated in Chapter 3 indeed seems to be a workable approach that permits the identification of abusive tax avoidance from a policy perspective. In addition, this Chapter has considered why Canadian courts have not applied the G A A R to three of the four transactions discussed. In that light, this Chapter identifies some significant failures in interpreting the G A A R . In particular, the courts sometimes fail to properly distinguish between the objective purposes of transactions and taxpayers' subjective goals in undertaking those transactions. Such a distinction is less important in the case of isolated tax avoidance transactions like the circular financing sale-leaseback in Canada Trustco, but it is of critical importance when tax-motivated steps are inserted in a series of transactions as they were in Evans and Canadian Pacific. Since an avoidance transaction is defined in terms of its purpose, a court absolutely must identify the purpose of each transaction appropriately i f there is to be any hope of applying the G A A R properly. 178 Furthermore, in addition to sometimes failing to properly apply the purpose test, the courts have failed to properly determine whether transactions result in a misuse or abuse (or, in the language of Canada Trustco, frustrate the intent of Parliament) because the courts do not consider the economic results, or outcomes, of transactions when applying subsection 245(4). I have argued that a proper understanding of Parliament's intent, or of the object and spirit of the provisions, inherently requires that the economic outcome of the transaction be considered because Parliament intends tax laws to apply tax consequences to transactions based on their economic outcomes. Although the underlying economic result of a transaction might not normally be relevant in determining the application of tax legislation (as emphasized in Shell), tax policy considerations require that, in the tax avoidance context, the G A A R be treated as a "provision to the contrary." Since Parliament presumes certain economic results to have occurred, it is only by considering whether taxpayers' transactions involve those economic results that the G A A R can be applied to ensure that normative ideal tax outcomes are reached. Courts that properly appreciate their role in the tax system, and the nature of tax law in general, w i l l treat the economic outcomes of taxpayers' transactions as a relevant consideration in determining whether to strike down tax avoidance transactions under the G A A R . However, the fact that Canadian courts are not doing so indicates that the courts do not share that appreciation. It may be that the best way to mandate this sort of analysis is by amending the G A A R to clarify that whether transactions result in a misuse or in an abuse is a determination to be made in light of their underlying economic substance. Parliament felt that a statutory G A A R was necessary to mandate a new approach to tax avoidance after Stubart demonstrated that Canadian courts were not prepared to assume responsibility for combating tax avoidance on their own. It appears that, despite the presence of the G A A R , Canadian courts are falling back into the old constraints of the Duke of Westminster approach. Perhaps only another statutory amendment mandating a fundamental departure from the old approach would spur them to apply the G A A R more vigorously. 180 VII . C O N C L U S I O N S In this thesis, I have argued that, from a tax policy perspective, tax avoidance is a significant problem. Tax avoidance is a significant policy problem because it frustrates the most important objectives of the tax system - the raising of revenue and the redistribution of income and wealth - and because it distorts important features of the tax system such as equity, neutrality, simplicity, and efficiency. Finally, tax avoidance poses a direct challenge to fairness and an equitable distribution of the tax burden, upon which rest public confidence and willingness to comply. Public confidence, in turn, underpins the integrity and credibility of our tax system, which are essential to a self-assessing system. I have argued that it is possible to identify certain tax reducing activities as abusive while still acknowledging that, in a market economy founded on private property, taxpayers ought to be permitted to plan their affairs with tax in mind and to engage in tax-minimizing activities to the extent that Parliament intends to permit them. From a tax policy perspective, therefore, tax avoidance occurs when taxpayers undertake transactions for the purpose of obtaining tax benefits that result when the tax consequences imposed by the legislation on the basis of valid legal relationships are not the tax consequences that Parliament intended, or would have intended, to apply based on the underlying economic outcomes of those transactions. This definition is premised on the view that tax laws are a policy instrument through which Parliament attempts to achieve broader fiscal and economic policies by imposing tax consequences on underlying economic circumstances and outcomes. In addition, this theory acknowledges that there is a public interest in a tax system that is as predictable and as certain as possible, and that taxpayers should be entitled to rely on 181 straightforward tax laws as much as possible when arranging their economic affairs. Some forms of tax reduction activities are legitimate examples of tax mitigation as opposed to abusive tax avoidance. Therefore, the theory formulated in Chapter 3 requires, first, that the transactions giving rise to tax avoidance be undertaken primarily or solely for the purpose of obtaining those tax benefits and not for any other purpose. Second, the theory requires that the divergence between the tax consequences resulting from taxpayers' legal relationships and the tax consequences intended to apply based on their economic circumstances not be intended by Parliament. In other words, tax avoidance exists only where taxpayers have not incurred the economic consequences that Parliament intended them to incur in order to qualify for a tax reduction, and where Parliament has not specifically encouraged tax-motivated transactions through tax incentive provisions. In addition, this thesis has reviewed some of the approaches to controlling tax avoidance taken in other jurisdictions, and has concluded that none of those approaches have been adopted in Canada. The approach to interpreting tax legislation laid down by the Supreme Court of Canada does not permit any meaningful extra-statutory anti-avoidance doctrines to flourish. Only the sham transaction and ineffective transaction doctrines, which are arguably not real anti-avoidance transactions at all , exist in Canadian tax law. Doctrines that have proved useful in other jurisdictions, such as the business purpose test, step transaction doctrine, economic substance test, and abuse of rights doctrine, have not been adopted in Canada. Nor, given the narrow Canadian approach to interpreting tax statutes, is there room for the development of meaningful anti-avoidance doctrines distinct from the G A A R . The review of the various anti-avoidance doctrines conducted in Chapter 4 leads to the inescapable conclusion that the G A A R is effectively the be-all and end-all of anti-avoidance in Canada. This thesis has also considered the structure of the G A A R and how it is to be applied. Concepts such as tax benefit, avoidance transaction, series of transactions, and misuse and abuse are distinct concepts within a unique statutory framework. However, those concepts draw on some of the more common approaches to identifying tax avoidance. For instance, the G A A R uses a purpose-oriented approach to identifying avoidance transactions and a policy-oriented approach to determining whether those avoidance transactions constitute abusive tax avoidance. This thesis endorses the comparative approach that Canadian courts have taken to identifying a tax benefit. In addition, this thesis endorses the approach taken to connecting transactions into a series for purposes of identifying an avoidance transaction. I have also concluded that the primary non-tax purpose test for an avoidance transaction is best interpreted as a test of the objective purpose of a transaction as opposed to a test of a taxpayer's subjective intention or motivation for undertaking the transaction. The objective approach is particularly justified by the fact that subsection 245(3) is clearly aimed at step transactions in which transactions with tax-reduction purposes are inserted into series of transactions that themselves are undertaken in pursuit of legitimate economic or commercial objectives. In Canada Trustco, the Supreme Court of Canada generally characterized the analysis of whether a transaction results in a misuse or abuse under subsection 245(4) as an analysis 4 7 4 1 would argue that, given the presence of the GAAR in the Act, the Canadian tax authorities are extremely unlikely to even attempt to argue an extra-statutory anti-avoidance doctrine, and the Courts are likely to reject the development of any such doctrines as moot given the introduction of a statutory general anti-avoidance rule. 183 of whether a transaction results in abusive tax avoidance. To the extent that Canada Trustco requires the analysis to focus on the object and spirit of the Act, as determined by reviewing the text, context, and purpose of its provisions in light of the legislative scheme, rather than by considering deeper tax policy questions relating to why Parliament has enacted the rules that it has, Canada Trustco is a welcome development. However, to the extent that Canada Trustco relegates considerations of economic substance, or the underlying economic outcomes of transactions, only to situations in which the relevant legislative provisions expressly make it relevant, the current Canadian approach to tax avoidance is inadequate. The current approach ignores the fact that Parliament generally intends the Act to apply to transactions based on their underlying economic circumstances. Finally, I have proceeded to test the theory of tax avoidance developed in Chapter 3 by reviewing some real Canadian tax avoidance cases. I have argued that, from a tax policy perspective, the Canada Trustco and McNichol/Evans transactions constituted abusive tax avoidance. In those cases, it is possible to identify a normative tax treatment that Parliament would have intended to apply to the outcome of those transactions. I have also argued that, despite the courts' decisions in Canada Trustco and Evans, the G A A R should apply to those transactions and should be interpreted in a way that ensures that the normative tax treatment is achieved. This thesis concludes that the courts did not properly apply the G A A R in Canada Trustco and Evans because the courts did not appreciate that the underlying economic outcomes of the transactions in those cases were relevant. In addition, the courts have not always applied the purpose test in subsection 245(3) properly due to confusion about the objective purpose of transactions and taxpayers' subjective motivations for undertaking them. However, it should be possible to apply the G A A R in a way that does 184 ensure normative tax outcomes by employing the theoretical understanding of tax avoidance developed in Chapter 3. The review of the Canada Trustco, McNichol, and Evans transactions suggests that such an application is indeed possible. As Arnold observes, the question of tax avoidance raises "fundamental issues about the relationship between a taxpayer and the state." 4 7 5 Identifying tax avoidance reflects a "tension" between the private property rights of taxpayers and the government's responsibility to preserve the tax base and a fair distribution of the tax burden for the benefit of all taxpayers. 4 7 6 Furthermore, unfettered tax avoidance undermines the shared conception of justice on which we base the distribution of social costs reflected in the tax system. To the extent that unchecked tax avoidance undermines public confidence in the integrity and fairness of the tax system, it has potentially damaging effects on social stability and democracy as well . Tax avoidance is a unique area of law, because it deals with activities that are legal but which, when considered from a public policy perspective, nevertheless should be controlled. Perhaps that tension is best summed up by Lord Simon: There is no doubt that they [i.e., tax avoiders] are within their legal rights, but that is no reason why their efforts or those of their professional advisers should be regarded as a commendable exercise of ingenuity or as a discharge of the duties of good citizenship 4 7 7 Fairness, justice and good citizenship are all implicated in the tax system, and tax avoidance is inconsistent with all of them. In order to protect those interests, Canadian tax law needs a rule like the G A A R , and that rule needs to be applied vigorously and meaningfully in order to give effect to Parliament's true intention. Arnold, "Canadian General Anti-Avoidance Rule", supra note 354 at 243. Arnold, "Reflections", supra note 10 at 32. Latilla v. Inland Revenue Commissioners, [1943] A.C. 377 at 318 (H.L.). B I B L I O G R A P H Y 185 L E G I S L A T I O N Bank Act, R .S .C . 1985, c. B - l . Income Tax Act, R .S .C . 1985, c. 1 (5 t h Supp.), as amended. Income Tax Act, 2004 (New Zealand). Income Tax Assessment Act 1936 (Australia). Income Tax Regulations, C R C c. 945, as amended. Partnership Act, R .S .O. 1980, c. 370. Tax Avoidance Act, SFS 1980:865 (Sweden). United States Revenue Act of 1928, c. 852, 45 Stat. 791. C A N A D I A N J U R I S P R U D E N C E 65302 B.C. Ltd. v. R., 99 D .T .C . 5799 (S.C.C.). 771225 Ontario Inc. v. Bramco Holdings Ltd. (1995), 21 O.R. (3d) 739 (Ont. C .A . ) , affirming 17 O.R. (3d) 571 (Ont. Gen. Div.) . Antosko v. R., 94 D .T .C . 6314 (S.C.C.). Atinco Paper Products Ltd. v. The Queen, 78 D .T .C . 6387 (F .C.A.) . Canada (Attorney General) v. Juliar, 2000 D .T .C . 6589 (Ont. C . A . ) , affirming 46 O.R. (3d) 104 (Ont. Sup. Ct.). Canada Trustco Mortgage Company v. R., 2005 D . T . C . 5523 (S.C.C.), affirming 2004 D .T .C . 6119 (F .C.A.) , affirming 2003 D .T .C . 587 (T.C.C.) . Continental Bank Leasing Corp. v. The Queen, 98 D .T .C . 6505 (S.C.C.), reversing 96 D .T .C . 6355. Continental Bank of Canada v. R., 94 D .T .C . 1858 (T.C.C.) . 186 Dale v. R., 97 D .T .C . 5252 (F.C.A.) . Duha Printers (Western) Ltd. v. The Queen, 98 D . T . C . 6334 (S.C.C.) . Evans v. R., 2005 D .T .C . 1762 (T.C.C.) . Friesen v. R., 95 D . T . C . 5551 (S.C.C.). Hill v. R., 2002 D .T .C . 1749 (T.C.C.) . Imperial Oil Ltd. v. The Queen, 2004 D . T . C . 6044 (F .C.A. ) , affirming 2002 D . T . C . 1954 (T.C.C.) . M.N.R. v. Leon et al, 76 D .T .C . 6299 (F .C.A.) . Massey-Ferguson Ltd. v. The Queen, 11 D . T . C . 5013 (F .C.A.) . Matthew v. Canada, 2005 D . T . C . 5538 (S.C.C.). McNichol v. The Queen, 91 D .T .C . I l l (T.C.C.) . Neuman v. The Queen, 98 D . T . C . 6297 (S.C.C.). Notre-Dame de Bonsecours v. Communaute Urbaine de Quebec et al, 95 D .T .C . 5017 (S.C.C.). OSFC Holdings Ltd. v. R., 2001 D .T .C . 5471 (F .C.A.) . R. v. Province of Alberta Treasury Branches and Toronto Dominion Bank, 96 D .T .C . 6245 (S.C.C.). Shell Canada Ltd. v. R., 99 D .T .C . 5669 (S.C.C.). Singleton v. R., 2001 D . T . C . 5533 (S.C.C.). Singleton v. R., 96 D . T . C . 1850 (T.C.C.) . Snow White Productions Inc. v. PMP Entertainment Inc. et al, 2005 D . T . C . 5150 (B.C.S.C.) . Stubart Investments Ltd. v. R, 84 D .T .C . 6305 (S.C.C.), reversing 81 D . T . C . 5120 (F.C.A.) . The Queen v. Bronfman Trust, 87 D . T . C . 5059 (S.C.C.). The Queen v. Canadian Pacific Ltd., 2002 D . T . C . 6742 (F .C.A.) . 187 Water's Edge Village Estates (Phase II) Ltd. v. R., 2002 D . T . C . 7172 (F .C.A.) . Will-Kare Paving & Contracting Ltd. v. R., 2000 D . T . C . 6467 (S.C.C.). U N I T E D K I N G D O M J U R I S P R U D E N C E Barclays Mercantile Business Finance Limited v. Mawson (Inspector of Taxes), [2004] U . K . H . L . 51 (H.L.) . J Challenge Corporation Ltd. v. Commissioner of Inland Revenue, [1987] A . C . 155 (P.C.). Craven v. White, [1989] A . C . 398 (H.L.) . Ensign Tankers (Leasing) Ltd. v. Stokes (Inspector of Taxes), [1992] 1 A . C . 655 (H.L.) . Furniss (Inspector of Taxes) v. Dawson, [1984] A . C . 474 (H.L.) . Inland Revenue Commissioners v. Brebner, [1967] 2 A . C . 18 (H.L.) . Inland Revenue Commissioners v. Burmah Oil Co. Ltd., [1982] S.T.C. 30 (H.L.) . Inland Revenue Commissioners v. Duke of Westminster, [1936] A . C . 1 (H.L.) . Inland Revenue Commissioners v. McGuckian, [1997] S.T.C. 908 (H.L.) . Inland Revenue Commissioners v. Scottish Provident Institution, [2004] U . K . H . L . 52 (H.L.) . Inland Revenue Commissioners v. Wesleyan & Gen. Assur. Soc. (1948), 30 Tax Cas. 11 (H.L.) . Inland Revenue Commissioners v. Willoughby, [1997] S.T.C. 995 (H.L.) . Latilla v. Inland Revenue Commissioners, [1943] A . C . 377 (H.L.) . MacNiven v. Westmoreland Investments Ltd., [2003] 1 A . C . 311 (H.L.) . Newton v. Commissioner of Taxation, [1958] A . C . 450 (P.C.). Partington v. Attorney-General (1869), L . R . 4 (H.L.) . Peterson v. Commissioners of Inland Revenue, [2005] U . K . P . C . (P.C.). Snook v. London & West Riding Investments Ltd., [1967] 1 A l l E .R. 518 (C.A.) . 188 W.T. Ramsay Ltd. v. I.R.C, [1982] A . C . 300 (H.L.) . U N I T E D S T A T E S J U R I S P R U D E N C E ACM Partnership v. Commissioner of Internal Revenue, 73 T . C . M . 2189 (1997). Commissioner of Internal Revenue v. Gordon, 391 U.S . 83. (1968). Commissioner of Internal Revenue v. Newman, 159 F.2d 848 (2d Cir . 1947). Compaq Computer Corp. v. Commissioner of Internal Revenue, 113 T . C . 214 (1999). DeMartino v. Commissioner, 862 F.2d 400 (2d Cir . 1988). Frank Lyon & Co. v. United States, 435 U.S . 561 (1978). Goldstein v. Commissioner of Internal Revenue, 364 F.2d 734 (2d Cir . 1966). Gregory v. Helvering, 293 U.S . 465 (1935), affirming Helvering v. Gregory, 69 F.2d 809 (2d Cir . 1934). Jacobson v. Commissioner, 915 F.2d 823 (2d Cir . 1990). Knetsch v. United States, 364 U.S . 361 (1960). Long Term Capital Holdings v. United States, 330 F. Supp. 2d 122 (D. Conn., 2004). Magneson v. Commissioner of Internal Revenue, 753 F.2d 1490 (9 t h Cir . 1985). McDonald's Restaurants of Illinois Inc. v. Commissioner of Internal Revenue, 76 T . C . 972 (1981). Redding v. Commissioner of Internal Revenue, 630 F.2d 1169 (7 t h Cir . 1980). UPS v. Commissioner of Internal Revenue, 245 F.3d 1014 (11 t h Cir . 2001). S E C O N D A R Y M A T E R I A L S ; B O O K S A N D M O N O G R A P H S 189 Boadway, Robin W . and Kitchen, Harry M . Canadian Tax Policy, 3 r d ed. Canadian Tax Paper No. 103 (Toronto: Canadian Tax Foundation, 1999). Canada. Report of the Royal Commission on Taxation, vol. 3, Appendix A (Ottawa: Queen's Printer for Canada, 1966). Dreidger, E . A . The Construction of Statutes, 2d ed. (Toronto: Butterworths, 1983). Duff, David. Canadian Income Tax Law (Toronto: Edmond Montgomery Publications Limited, 2003). Hogg, Peter W. , Magee, Joanne and L i , Jinyan. Principles of Canadian Tax Law, 4 t h ed. (Toronto: Carswell, 2002). Krishna, Vern. Fundamentals of Canadian Income Tax, 8 t h ed. (Toronto: Carswell, 2004). Tax Law Review Committee. Tax Avoidance (London: The Institute for Fiscal Studies, 1997). Ontario. Fair Taxation in a Changing World: Report of the Ontario Fair Tax Commission (Toronto: Queen's Printer for Ontario, 1993). Parsons, R . W . Income Taxation in Australia (Sydney: Law Book Co. , 1985). Surrey, Stanley S. and McDaniel , Paul R. Tax Expenditures (Cambridge, Mass.: Harvard University Press, 1985). S E C O N D A R Y M A T E R I A L S : A R T I C L E S Angell , Montgomery B . "Tax Evasion and Tax Avoidance" (1938) Colum. L . Rev. 80. Arnold, Brian. "In Praise of the Business Purpose Test" in Report of the Proceedings of the Thirty-Ninth Tax Conference, 1987 Conference Report (Toronto: Canadian Tax Foundation, 1988) 10:1. . "The Canadian General Anti-Avoidance Rule" in Graeme S. Cooper, ed. Tax Avoidance and the Rule of Law (Amsterdam: I B F D Publications B V , 1997) 221. -. "Reflections on the Relationship Between Statutory Interpretation and Tax Avoidance" (2001) 49:1 Can. Tax J. 1. 190 . "The Long, Slow, Steady Demise of the General Anti-Avoidance Rule" (2004) 52:2 Can. Tax J. 488. Arnold, Brian and Wilson, James. "The General Anti-Avoidance Rule - Part I" (1988) 36:4 Can. Tax J. 829. . "The General Anti-Avoidance Rule - Part II" (1988) 36:5 Can. Tax J. 1123. . "The General Anti-Avoidance Rule - Part III" (1988) 36:6 Can. Tax J. 1369. Baker, Philip. "Tax Avoidance, Tax Mitigation and Tax Evasion," online: Gray's Inn Tax Chambers, London . Bankman, Joseph. "The Economic Substance Doctrine" (2000) 74 S. Ca l . L . Rev. 5. Brooks, Michael and Head, John. "Tax Avoidance: In Economics, Law and Public Choice" in Graeme S. Cooper, ed. Tax Avoidance and the Rule of Law (Amsterdam: I B F D Publications B V , 1997) 53. Brooks, Ne i l . "The Changing Structure of the Canadian Tax System: Accommodating the R i c h " (1993) 31 Osgoode Hal l L . J . 137. . "The Responsibility of Judges in Interpreting Tax Legislation" in Graeme S. Cooper, ed. Tax Avoidance and the Rule of Law (Amsterdam: I B F D Publications B V , 1997) 93. Cooper, Graeme S. "Conflicts, Challenges and Choices - The Rule of Law and Ant i -Avoidance Rules" in Graeme S. Cooper, ed. Tax Avoidance and the Rule of Law (Amsterdam: I B F D Publications B V , 1997) 1. Couzin, Robert. "Subsection 245(3): A Framework" in Report of the Proceedings of the Forty-Ninth Tax Conference, 1997 Conference Report (Toronto: Canadian Tax Foundation, 1998) 4:1. Dodge, David A . " A New and More Coherent Approach to Tax Avoidance" (1988) 34:1 Can. Tax J. 1. Duff, David. "The Supreme Court of Canada and the General Anti-Avoidance Rule: Canada Trustco and Matthew," prepared for the University of Toronto G A A R Symposium November 18, 2005. Erlichman, Harry et al. "The Statutory Context of the G A A R " in Harry Erlichman, ed. Tax Avoidance in Canada (Toronto: Irwin Law Inc., 2002) 5. 191 Felesky, Brian A . and Jack, Sandra E . "Is There Substance to Substance Over Form in Canada?" in Report of the Proceedings of the Forty-Fourth Tax Conference, 1992 Conference Report (Toronto: Canadian Tax Foundation, 1993) 50:1. Fortin, Guy. "Abuse or Misuse" in Report of the Proceedings of the Forty-Ninth Tax Conference, 1997 Conference Report (Toronto: Canadian Tax Foundation, 1998) 5:1. Freedman, Judith. "Defining Taxpayer Responsibility: In Support of a General Ant i -Avoidance Principle" [2004], no. 4 British Tax Review 332. . "Converging Tracks? Recent Developments in Canadian and U K Approaches to Tax Avoidance" (2005) 53:4 Can. Tax J. 1038. Gammie, Malcolm. "Tax Avoidance and the Rule of Law: A Perspective from the United Kingdom" in Graeme S. Cooper, ed. Tax Avoidance and the Rule of Law (Amsterdam: I B F D Publications B V , 1997) 181. . "Barclays and Canada Trustco: Further Comment from a U K Perspective" (2005) 53:4 Can. Tax J. 1047. Gideon, Kenneth W . and Kent, Ruth E . "Mrs. Gregory's Northern Tour: Canadian Proposals to Adopt a Business Purpose Test" in Report of the Proceedings of the Thirty-Ninth Tax Conference, 1987 Conference Report (Toronto: Canadian Tax Foundation, 1988)7:1. Gunn, Alan. "Tax Avoidance" (1979) 76:5 Mich . L . Rev. 733. Hariton, David P. "Sorting Out the Tangle of Economic Substance" (1998-1999) 52:2 Tax Lawyer 235. Harvey, Brian W . "Tax Avoidance - Illegal, Immoral or Fattening?" (1970) 21 N . Ir. Legal Q. 235. Hoffman, Leonard. "Tax Avoidance" [2005], no. 2 British Tax Review 197. Isenbergh, Joseph. "Musings on Form and Substance in Taxation" (1982) 49 U . Chi . L . Rev. 859. Kay, J .A. "The Economics of Tax Avoidance" [1979], no. 6 British Tax Review 354. Keinan, Yoram. "The Many Faces of the Economic Substance Doctrine's Two-Prong Test: Time for Reconciliation?" (2005) 1 N . Y . U . J . L . & Bus. 371. Kellough, Howard J. " A Review and Analysis of the Redrafted General Anti-Avoidance Rule" (1988) 36:1 Can. Tax J. 23. 192 L i , Jinyan. "The Economic Substance Doctrine and G A A R : A Critical and Comparative Perspective," written for the University of Toronto G A A R Symposium November 18, 2005. McDonnel l , Thomas E . "The Evans Case: G A A R in a Post-Canada-Trustco Wor ld" (2006) 6:1 Tax for the Owner-Manager 9. Muten, Leif. "The Swedish Experiment with a General Anti-Avoidance Rule" in Graeme S. Cooper, ed. Tax Avoidance and the Rule of Law (Amsterdam: I B F D Publications B V , 1997) 307. Patton, Stewart. "Treasury Regulation § 301.61112T and the Economic Substance Doctrine: A Plea for Certainty in the Tax L a w " (2002-2003) 39 Hous. L . Rev. 499. Philipps, Lisa. "The Supreme Court of Canada's Tax Jurisprudence: What's Wrong With The Rule of L a w " (2000) 79 Can. Bar Rev. 120. Rosenberg, Joshua D . "Tax Avoidance and Income Measurement" (1988-1989) 87 M i c h . L . Rev. 365. Sawyer, Adrian J. "Blurring the Distinction Between Avoidance and Evasion - The Abusive Tax Position" (1996) 5 British Tax Review 483. Taylor, Roger. "The Supreme Court of Canada: Principles of Adjudication of Tax-Avoidance Appeals from Stubart to Shell Canada" in Report of the Proceedings of the Fifty-First Tax Conference, 1999 Conference Report (Toronto: Canadian Tax Foundation, 2000) 17:1. Tiley, John. "Judicial Anti-Avoidance Doctrines: The U S Alternatives" [1987], no. 5 British Tax Review 180. . "Judicial Anti-Avoidance Doctrines: The U S Alternatives - Part II" [1987], 6 British Tax Review 220. . "Judicial Anti-Avoidance Doctrines: Some Problem Areas" [1988], no. 3 British Tax Review 63. . "Judicial Anti-Avoidance Doctrines: Corporations and Conclusions" [1988], no. 4 British Tax Review 108. . "Tax Avoidance Jurisprudence as Normal L a w " [2004], no. 4 British Tax Review 304. Tiley, John and Jensen, Erik. "The Control of Avoidance: The United States Experience" [1998], no. 2 British Tax Review 161. 193 Troup, Edward. "Unacceptable Discretion: Countering Tax Avoidance and Preserving the Rights of the Individual" (1992) 13:4 Fiscal Studies 128. Uckmar, V . "Tax Avoidance/Tax Evasion" (1983) 68a Cahiers de droit fiscal international 15. Vanistendael, Frans. "Judicial Interpretation and the Role of Anti-Abuse Provisions in Tax L a w " in Graeme S. Cooper, ed. Tax Avoidance and the Rule of Law (Amsterdam: I B F D Publications B V , 1997) 131. Waincymer, Jeffrey. "The Australian Tax Avoidance Experience and Responses: A Critical Review," in Graeme S. Cooper, ed. Tax Avoidance and the Rule of Law (Amsterdam: I B F D Publications B V , 1997) 247. Ward, David. "Tax Avoidance: Judicial and Legislative Approaches in Other Jurisdictions" in Report of the Proceedings of the Thirty-Ninth Tax Conference, 1987 Conference Report (Toronto: Canadian Tax Foundation, 1988) 8:1. Ward, et al. "The Business Purpose Test and Abuse of Rights" [1985], no. 1 British Tax Review 68. Wilk ie , Scott and Kerr, Heather. "Common Links Among Jurisdictions: Informing the G A A R Through Comparative Analysis" in Report of the Forty-Ninth Tax Conference, 1997 Conference Report (Toronto: Canadian Tax Foundation, 1998) 34:1. O T H E R M A T E R I A L S Concise Oxford Dictionary, 9 t h ed. (Oxford: Clarendon Press, 1995). Department of Finance Canada. Technical Notes, 17 t h ed. Ed. , David M . Sherman (Toronto: Thomson Canada Limited, 2005.)