If It Ain’t Broke, Don’t Fix It: Examining the Amendments to the General Anti-Avoidance Rule
Nathan McLean, 3L JD/MBA, Senior Editor
The GAAR and the Proposed Amendments
In 2023, the Canadian federal government proposed significant amendments to the general anti-avoidance rule (GAAR) of the Income Tax Act (ITA), a cornerstone in combatting abusive tax planning (Budget 2023, Tax Measures: Supplementary Information). Originating in 1988, the GAAR addresses tax avoidance policy concealed by technical compliance. The proposed changes include a new preamble, an altered definition for “avoidance transaction,” an economic substance test, and a penalty equal to 25% of the tax benefit claimed (with some exceptions). These changes mark a transformative shift. I will examine some of the key proposed changes to the GAAR, arguing that the amendments unnecessarily complicate the ITA and lead to more confusion than they are worth.
In order to understand the function of the GAAR in Canada’s tax system, it is important to understand the general analytical principles used in assessing potential tax abuse cases. The traditional approach in the Canadian case law rejects the economic substance doctrine—which focuses on the economic value of a transaction—and instead considers the legal form of transactions. This sharply contrasts with the dominant approach in the US, where the substance over form approach allows tax authorities to deny certain benefits where the form and substance of a transaction differ such that it lacks a valid business purpose.
Traditionally, the Canadian approach followed the guiding principle from the seminal case Inland Revenue Commissioners v Duke of Westminster, [1936] AC 1 (Duke of Westminster). The Court placed a paramount focus on a taxpayer’s actual legal relationships, steering clear of delving into the intricacies of economic realities or overall business purposes. The echoes of Duke of Westminster reverberate in contemporary Canadian cases, such as Singleton v Canada, 2001 SCC 61, where the Supreme Court of Canada (SCC) held that a “shuffle of cheques” that defines the legal relationship must be given effect. In essence, the GAAR counterbalances the paramountcy of this “shuffle of cheques.” If an “avoidance transaction” (or series of transactions) is abusive, the GAAR will deny tax benefits incurred while violating the object, spirit, and purpose of the provisions that gave rise to the benefit. In this way, the GAAR “attenuate[s]” the form over substance doctrine that distinguishes the Canadian approach; Deans Knight Income Corp v Canada, 2023 SCC 16 at para 47.
At its core, the GAAR was legislated “to distinguish between legitimate tax planning and abusive tax avoidance and to establish a reasonable balance between the protection of the tax base and the need for certainty for taxpayers in planning their affairs”; Department of Finance Canada, Explanatory Notes to Legislation Relating to Income Tax (1988) at 461. Unfortunately, the proposed changes to the GAAR tilt the scales, undermining the simplicity and predictability of the tax system.
The Sectional Preamble
As part of the GAAR amendments, Parliament will include a sectional preamble in s. 245(0.1) of the ITA. The preamble explains the purpose of the GAAR and attempts to define its scope. There are a few key aspects worth noting.
The proposed amendment in s. 245(0.1)(a) permits taxpayers to obtain tax benefits as envisaged by the relevant provisions of the ITA. However, the language used in this subsection of the preamble (“contemplated by the relevant provisions”) introduces a layer of complexity by requiring taxpayers to discern and understand Parliament’s intentions when legislating specific provisions. Does this imply that taxpayers must delve into the legislative history, policy objectives, and the overall purpose behind each provision to navigate the contours of permissible tax benefits? The obligation for taxpayers to interpret and align their actions with Parliament’s contemplation generates uncertainty and gives rise to potential disputes where Parliament’s intentions are unclear.
In the proposed amendments to the GAAR under s. 245(0.1)(b), the explicit inclusion of the term “fairness” in the sectional preamble introduces a nuanced dimension to the government’s objectives. While one might initially interpret the inclusion of fairness as an attempt to underscore the equity advantages inherent in the avoidance rule, a closer examination raises legitimate concerns about its potential implications. The lack of clarity surrounding whether fairness is intended to function as a novel interpretive principle for applying the GAAR introduces ambiguity. This ambiguity could give rise to divergent interpretations and disputes, as fairness lacks a precise and universally accepted definition. Without clear guidance on how fairness is to be defined and applied, taxpayers may find themselves navigating uncertain terrain when assessing the legitimacy of their tax planning strategies, potentially leading to increased challenges and conflicts in the interpretation and application of the GAAR.
The addition of fairness as a third objective in s. 245(0.1)(b)(ii) raises intriguing questions about the original equilibrium sought by the GAAR. As noted above, the initial stated objective of the GAAR was to strike a delicate balance between safeguarding the tax base and providing certainty for taxpayers in their tax planning activities. The introduction of fairness as a separate factor raises questions about whether this new objective is inherent within the protection of the tax base or if it introduces a distinct dimension. If the concept of fairness is considered integral to the safeguarding of the tax base, why explicitly include it as a separate goal? Alternatively, if fairness introduces an independent consideration, it invites a re-evaluation of the established equilibrium and poses the challenge of how these three objectives interact in practice. If this is the case, it risks adding further complexities that may disrupt the delicate equilibrium originally sought by the rule.
The overarching role of the sectional preamble remains ambiguous within the context of the proposed amendments to the GAAR. It raises a fundamental question: is the preamble intended to reshape the GAAR’s core principles, or does it primarily serve as a vehicle for introducing the forthcoming amendments and the new GAAR? The fact that this question remains unanswered underscores the complexity added by the amendments.
The Decreased Burden
Before the amendments, the GAAR test considered “whether the transaction is an avoidance transaction under s. 245(3), in the sense of not being ‘arranged primarily for bona fide purposes other than to obtain the tax benefit’”; Canada Trustco Mortgage Co v Canada, 2005 SCC 54 at para 17. In the proposed amendments to s. 245(3)(b), courts will now consider whether “one of the main purposes for undertaking or arranging the transaction was to obtain the tax benefit” in order to deem it an “avoidance transaction” under the ITA.
Shifting the focus from a transaction being primarily arranged for purposes other than tax benefits to whether obtaining the tax benefit was one of the main purposes raises concerns about potentially expanding the scope of “avoidance transaction” without justification. This is especially the case given that the original threshold for categorizing a transaction as an “avoidance transaction” under s. 245(3) was not particularly burdensome for tax authorities to meet. The proposed amendments risk broadening the scope to a threshold that may encompass routine and benign tax planning activities. The proposed change could include transactions not initially intended to fall under GAAR, impacting a wide array of taxpayer activities. Without a clear rationale, the decreased standard risks eroding the balance between preventing abusive tax practices and allowing legitimate tax planning.
Consider a hypothetical (and simplified) scenario where a corporation restructures its subsidiaries to enhance operational efficiency, streamline business operations, and reduce its tax burden. Under the current framework, if these bona fide non-tax purposes are deemed the primary motivation, the transaction would be acceptable—this type of activity should not elicit accusations of being an avoidance transaction. However, under the proposed amendment, if one of the main purposes of the restructuring is also to obtain a tax benefit, this transaction could be considered an avoidance transaction. The broader criterion risks scrutinizing legitimate business decisions that involve tax considerations as one of their main purposes, potentially bringing routine activities under the GAAR’s purview. Such an expansion of the GAAR's reach could create a chilling effect on corporate decision-making, as businesses may become more cautious about engaging in transactions that have a legitimate mix of bona fide business and tax-related objectives for fear of the potential characterization as avoidance transactions.
Economic Substance and the Reverse Onus
Before the proposed amendments, the GAAR test burdened taxpayers with the onus of (i) refuting the presumption that they had received a benefit, and (ii) that the benefit was the product of an avoidance transaction. However, the onus was on the Crown to demonstrate that the transaction resulted in misuse or abuse. Under the proposed amendments, there is a presumption that avoidance transactions “significantly lacking in economic substance” result in a misuse or abuse. Now, the onus is on the taxpayer to demonstrate that the transaction is not a misuse or abuse.
At a basic level, the reverse onus risks overhauling a fundamental principle of Canadian tax law which is generally based on legal form rather than economic substance. Instead, courts must now weigh the expected value of the tax benefit against the expected non-tax economic return and analyze the business determinations that companies make to determine the purpose for undertaking the transaction. This goes directly against the current approach in the case law; see Canada Trustco Mortgage Co. v Canada, 2005 SCC 54 at para 57.
Such granular investigation into the economic rationale for business decisions rejects the Duke of Westminster principle in favour of something that looks much more like an economic substance test. On the one hand, the existing GAAR already puts a leash on the form over substance doctrine. Taxpayers can only order their affairs to obtain a tax benefit when it is not abusive tax avoidance. However, the amendments, read as they are written, limit taxpayers’ ability to arrange their legal affairs to situations where the expected non-tax economic returns exceed the expected value of the tax benefit.
Despite the apparent move towards an economic substance test, the amended GAAR maintains an assertion of prioritizing the legal form of transactions. This duality creates a framework where the system claims to adhere to the legal form but also explicitly factors in the economic substance and expected value of transactions. This ambiguity risks confusing taxpayers, as the system attempts to straddle both legal formalities and economic realities.
Is the GAAR broken?
Between 1988 and March 2021, the Canada Revenue Agency considered the GAAR in 1,600 audit cases and applied it in 80% of cases; see Department of Finance Canada, “Modernizing and Strengthening the General Anti-Avoidance Rule.” The Crown has won all but two of the six GAAR cases heard by the SCC, including the most recent case in Deans Knight Income Corporation v Canada, 2023 SCC 16. This track record raises questions as to whether or not there were issues with the existing GAAR.
This is not to suggest that a broader tax avoidance problem does not exist. Yet, there are multiple possible sources for this outside of an ostensibly weak GAAR. In his recent article “AI and the Future of Tax Avoidance,” Professor Benjamin Alarie comments on the potential improvements artificial intelligence could yield in, among other things, detecting instances of tax avoidance; see page 1812. It is worth examining potential shortcomings in identifying tax avoidance activities. It is also worth considering the possibility that the Crown might be selective or cautious in applying the GAAR in certain cases. This selectivity could create gaps in enforcement and contribute to the perception that the GAAR is itself inadequate; see the Appendix in Paletta Estate v The Queen, 2021 TCC 11.
However, most notably, the government could consider the possibility that the GAAR works adequately to deter taxpayers from even engaging in abusive avoidance transactions to begin with. If the existing framework proves effective in dissuading individuals and corporations from participating in abusive avoidance transactions, it could naturally result in a lower number of cases. In this scenario, the focus might shift from solely assessing the GAAR’s application to a more holistic evaluation of its preventive impact on tax avoidance practices.
As Parliament previously acknowledged, it is essential to strike a balance between protecting the tax base by preventing abusive tax practices and affording taxpayers certainty in planning their affairs. Any alteration to the GAAR should ideally start from the premise that Parliament deemed that a prior case, where the GAAR was not applied, was incorrectly decided. After a retrospective analysis of the SCC’s GAAR jurisprudence, it is unclear whether the outcome of a single case would have changed on the basis of the revisions. Ultimately, the amendments complicate Canadian tax law without clear prospects of substantially reducing tax avoidance and safeguarding Canada’s coffers.