When Tax Avoidance is Abusive: Elucidating the “Spirit” of the Income Tax Act in Copthorne Holdings Ltd v Canada
When deciding cases relating to tax planning – or the minimization of a taxpayer’s tax burden – judges face the daunting task of reconciling a tension between longstanding common law principles and s. 245 of the Income Tax Act, RSC 1985, c.1, the General Anti-Avoidance Rule (GAAR). On the one hand, Lord Tomlin’s holding in Duke of Westminster that “every man is entitled if he can to order his affairs so as that the tax attaching under the appropriate Acts is less than it otherwise would be” is deeply entrenched in Canadian law. On the other hand, the enactment of the GAAR scheme was intended to impose constraints on this permissive common law doctrine by attaching tax liability to transactions structured deliberately to avoid tax.
What is clear in the caselaw is that the Duke of Westminster principle still holds in Canadian tax law (see Canada Trustco Mortgage Co v Canada,  2 SCR 601 [Canada Trustco Mortgage). What is unclear, however, is how, when, and to what degree the GAAR attenuates this principle. Recently, the Supreme Court of Canada (SCC) explored this tension in Copthorne Holdings Ltd. v Canada, 2011 SCC 63, refining the interpretive approach to GAAR but at the same time affirming the inescapable truth that “it is relatively straightforward to set out the GAAR scheme. It is much more difficult to apply it.”
The case involved a group of Canadian and non-resident companies controlled by Li Ka-Shing and his son, Victor Li. Through a series of transactions, two Canadian corporations within this group that had previously been parent and subsidiary became “sister” corporations, or corporations owned directly by the same non-resident shareholder. The nature of the amalgamation becomes relevant because of the tax treatment of paid-up capital (PUC), the capital invested in a class of shares of a corporation by its shareholders.
When that class of shares is redeemed by the corporation, the amount paid to shareholders includes an amount which is simply the return of capital (i.e. after-tax dollars invested in a corporation). This amount is not included in income as it has already been taxed. If the amount paid to shareholders exceeds the PUC (i.e. the amount invested), the excess amount is included in income as a dividend. It is in the interest of taxpayers, therefore, to structure their affairs so as to maximize the proportion of the payment which is attributable to PUC and thereby minimize the amount that is deemed to be a taxable dividend. This is exactly what Li Ka-Shing tried to do.
When corporations amalgamate, the general rule is that the PUC of the shares of both amalgamating corporations are aggregated to form the PUC of the new corporation. This does not apply, however, to an amalgamation involving a parent and subsidiary company (i.e. a vertical amalgamation). Upon amalgamation, the PUC of the shares of the subsidiary corporation owned by the parent company are cancelled to prevent the artificial inflation of PUC for tax purposes. Had the two Li companies remained as parent and subsidiary, the PUC of the shares of the subsidiary would have been cancelled on amalgamation. Since the two corporations had become sister corporations, the PUC was not cancelled and the amalgamated corporation proceeded to redeem a large portion of its shares to the non-resident shareholder.
The payment was treated by the taxpayer as a return of capital and thus not taxable income. Despite the fact that there is no specific provision in the Act requiring inclusion of the payment in income, the Minister of National Revenue deemed the transactions abusive within the meaning of GAAR and concluded that a deemed dividend arose on the redemption of the shares amounting to $58,325,223. The re-assessment was upheld by the Tax Court and the Federal Court of Appeal.
The Legal Context and the SCC Judgment
At the SCC, Rothstein J. affirmed the analytic framework for applying the GAAR set out in Canada Trustco Mortgage. Courts must ask three questions: (1) was there a tax benefit? (2) was the transaction giving rise to the tax benefit an avoidance transaction? (3) was the avoidance transactions giving rise to the tax benefit abusive? What is unusual about the GAAR is that while a taxpayer’s transactions will be in compliance with the text of the relevant provisions of the Act, they may contravene their object, spirit, or purpose. Thus, unlike traditional statutory interpretation which draws on the spirit of legislative provisions in order to make clear the meaning of the text, the GAAR interpretive approach requires courts to enforce not only the text of the Act but also the underlying rationale as it may extend beyond the text.
In other words, the GAAR, by definition, constitutes a legislative override of a canon of statutory interpretation often invoked by courts in deciding tax cases: if the legislature had intended to include something in a statutory provision, it would have done so and it is not for courts to judicially supplement the legislative process. In the context of GAAR, the starting premise of judicial analysis is that the rationale of a provision will not always be exhausted by the precise words of a statute.
In applying the three-part test outlined above, the SCC found that the transactions resulted in a tax benefit. While Rothstein J. accepted that the transactions were necessary to achieve a number of beneficial outcomes sought (e.g. simplification of the corporate structure and the ability to shelter anticipated gains with losses within the amalgamating corporations), the decision to opt for a horizontal instead of a vertical amalgamation produced a discernable benefit to the taxpayer.
Second, the SCC found that the transaction giving rise to the tax benefit was an avoidance transaction since it was part of the same series of transactions as the prior amalgamation and a vertical amalgamation would have resulted in the same non-tax benefits as the horizontal amalgamation undertaken by the taxpayers.
Finally, the transaction giving rise to the tax benefit was abusive. Generally, the SCC held that the case for abusive tax avoidance will be made out where: (1) the transaction achieves an outcome the statutory provision was intended to prevent; (2) the transaction defeats the underlying rationale of the provision; or (3) the transaction circumvents the provision in a manner that frustrates or defeats its object, spirit, or purpose. Rothstein J. found that the purpose of the provision was to preclude corporations from preserving PUC of the shares of a subsidiary corporation on amalgamation with the parent corporation as that PUC reflects the investment of the same tax-paid dollars in the subsidiary as in the parent corporation. Copthorne’s transactions, therefore, contravened this purpose.
Implications: Certainty, Flexibility, and the “Damoclesian Menace of the GAAR”?
A corollary of vesting courts with the power to enforce not only the statutory text of the Act but also its underlying rationale is a requirement that courts look deeper into corporate transactions and assess their economic realities. Traditionally, the common law approach of the courts has been to accept the legal characterization of transactions despite their economic realities if there is no “sham” or “window dressing.” Thus, what is most significant about GAAR is not so much the requirement that courts elucidate the spirit, object, or purpose of statutory provisions and give them concrete weight, but the corresponding obligation on courts to depart from their usual hesitancy to inquire into the economic realities of transactions. Through the GAAR, the legislature recalibrates the traditional judicial balance struck between individual liberty and equity.
The outcome of the decision is consistent with several underlying principles of the tax regime, particularly the need to tax people on their ability to pay and the use of the tax system to achieve horizontal and vertical equity. One of Copthorne’s arguments was that the current approach to the GAAR, in affording a great deal of leeway and flexibility for reassessments in light of the rationale of provisions, undermines another important goal of the tax system: certainty and predictability. Copthorne argued that upholding the Tax Court’s decision amounts to subjecting taxpayers to the “Damoclesian menace of the GAAR.”
The SCC was satisfied that there are adequate safeguards embedded in the analytic framework set out for GAAR to manage effectively the inevitably greater uncertainty. The Court pointed out that GAAR should only be turned to as a last resort and that in order for it to be triggered, there must be an avoidance transaction resulting in a tax benefit. It also highlighted the fact that if a transaction had a secondary tax benefit purpose, it would not engage the GAAR. Furthermore, the imposition of the burden on the Minister “who wishes to overcome the countervailing obligations of consistency and predictability to demonstrate clearly the abuse he alleges” protects taxpayers from the ‘menace’ of the GAAR.
Despite these safeguards, however, in the end the inescapable truth remains: it is relatively straightforward to set out the GAAR scheme. It is much more difficult to apply it.