Last August, the government released a discussion paper (the Discussion Paper) identifying a number of issues the government has with the general anti-avoidance rule (GAAR) in s. 245 of the Income Tax Act (Canada) (ITA) and suggesting various possible legislative amendments in order to expand its scope. The federal budget released on March 28, 2023 included a number of proposed amendments to s. 245 ITA (the March 2023 Proposals), which it summarized as follows:
Budget 2023 proposes to amend the GAAR by: introducing a preamble; changing the avoidance transaction standard; introducing an economic substance rule; introducing a penalty; and extending the reassessment period in certain circumstances.
Interested parties were invited to submit comments to the Department of Finance Canada, Tax Policy Branch at [email protected] by May 31, 2023, a deadline that is fast approaching. The Budget states that “following this period of consultation, the government intends to publish revised legislative proposals and announce the application date of the amendments.”
The source of the government’s apparent urgency is unclear. The inclusion of legislative proposals in the March 2023 budget was surprising to many in the tax community who believed the government would wait until the Supreme Court of Canada (SCC) has released its pending decision in the taxpayer’s appeal in Deans Knight Income Corporation v. The King, which will occur on May 26. However, the accelerated timeline that the government has adopted leaves the tax and business communities very little time to understand the potential scope of these changes to the most powerful provision in the ITA and provide feedback to the government on possible concerns.
The government has been very transparent with its objective in this exercise: it wants to make it easier for the Canada Revenue Agency (CRA) to successfully apply GAAR more often. In the materials it has released to date, the government’s reasons for so doing can be summarized within two principal themes:
- the existing text of GAAR is deficient from a policy perspective, in terms of (1) using a standard for “avoidance transaction” that is too high and thereby allowing abusive tax planning to occur, (2) not creating an adequate deterrent to abusive tax planning (hence the proposed penalty), and (3) not allowing a long enough re-assessment period; and
- the GAAR jurisprudence produced by the courts to date is not dealing appropriately with (1) the concept of “economic substance,” (2) the balance between protecting the tax base from abusive planning (styled as “fairness”) and the desire for giving taxpayers reasonable “certainty” in arranging their affairs, (3) applying GAAR to tax planning that was “foreseeable” by Parliament, and (4) the original intent of Parliament as expressed in the extrinsic aids that accompanied its enactment in 1988 (e.g., Department of Finance Technical Notes).
However, a careful review of the GAAR jurisprudence on these topics and the relevant extrinsic aids from 1988 reveal very little on which to rest these two premises. In neither the Discussion Paper nor the March 2023 Proposals does the government specify which GAAR cases it lost that it believes it should have won (and more importantly, why it should have won them), or which ones would be decided differently if the amendments being proposed were applicable. In doing so, the government overlooks the more readily-apparent conclusion from the GAAR jurisprudence (reinforced by the fact that the government only wins about half of the GAAR cases it litigates): the CRA loses GAAR cases because it seeks to apply GAAR aggressively in circumstances where it ought not to apply, and/or because the government has not sufficiently articulated the “legislative rationale”1 it claimed the taxpayer was contravening. The March 2023 Proposals address very little about these shortcomings and in fact their most likely impact will be to cause GAAR to be applied more frequently in cases where no demonstrable abuse or misuse is occurring while simultaneously undermining the guidance provided by 35 years of GAAR jurisprudence to date.
In its submission on the March 2023 Proposals, the Canadian Chamber of Commerce identified these concerns and reviewed the GAAR jurisprudence in detail, while reiterating the important role that GAAR plays in preventing the few who engage in abusive tax planning from shifting the burden of paying their fair share onto the many who are compliant and want only to remain on the right side of the line while paying no more than what Parliament intends. Both the government and the business community have a shared interest in ensuring that GAAR catches those few, and only those few. The objective of any exercise to enhance s. 245 ITA should be a GAAR that is carefully targeted at identifying and preventing tax planning that achieves results clearly contrary to Parliament’s evident intent, along with measures to ensure that the CRA administers GAAR in practice consistently with that common objective.
The practical impact of the March 2023 amendment to the “avoidance transaction” definition will be to make it largely irrelevant going forward, especially where tax advice has been obtained (it is already rarely an impediment to the CRA applying GAAR). This makes the abuse-or-misuse element of GAAR critically important and in this regard, any effort to reduce the government’s obligation to clearly establish the legislative rationale it claims the taxpayer is transgressing must be viewed very warily. It is essential for the government to articulate exactly what it is trying to achieve with the legislative amendments it has proposed. They appear superficially unobjectionable: for example, who can argue with describing GAAR as a balance between certainty and protecting the tax base? However, they are phrased in vague terms and the fact that the government has not identified which GAAR cases it thinks were wrongly decided and which ones its proposed amendments would change the results of raises the question: what is it that the government is trying to achieve with these changes? Having been repeatedly invited to answer these questions, if the government does not do so while proceeding with these vaguely-worded legislative changes to s. 245, a court considering their impact could reasonably infer that the government does not intend to materially change the existing GAAR guidance provided by the jurisprudence to date: surely significant changes in tax policy would be explicitly identified as such, as with the amendment to the “avoidance transaction” definition.
Such a conclusion would also be consistent with a careful review of the GAAR caselaw and the extrinsic aids from 1988. With respect to the proposed interpretational pre-amble, the GAAR caselaw:
- correctly interprets Parliament as placing the onus on the CRA to establish that for GAAR to be applicable, it must show the taxpayer’s actions to frustrating or defeating the relevant legislative rationale or achieving an outcome demonstrably contrary to Parliament’s intent (i.e. there is no reverse abuse-or-misuse onus on the taxpayer);
- already recognizes and considers the importance of protecting the tax base (whether or not termed “fairness”) in undertaking an abuse-or-misuse analysis; and
- does not indicate that courts will refuse to find an abuse or misuse simply because a taxpayer’s actions were foreseeable (although that may be a relevant data point).
Similarly, the caselaw clearly demonstrates that the courts are frequently willing to consider economic substance as relevant both to establishing the applicable legislative rationale and (separately) measuring the taxpayer’s actions against that standard where the ITA provisions in question make considering economic substance appropriate and pertinent. As such, without an unequivocal statement from the government that the intended effect of the March 2023 Proposals dealing with abuse or misuse demonstrably constitute a conscious change in tax policy (and clearly articulating the scope of that change), a court interpreting them might reasonably conclude that they simply entrench legislatively what the courts are already doing in this regard, so as to eliminate doubt and the risk of variation and misinterpretation.
The proposed 25 per cent penalty where GAAR applies (subject to relief where notification is made in prescribed form) seems difficult to justify, both as a matter of tax policy and as a matter of basic fairness. The Discussion Paper cites only one example of a court declining to apply a penalty in a GAAR case, the ITA already includes various penalty provisions potentially applicable in appropriate circumstances (e.g., s. 163(2)), the idea of a penalty applying no matter how reasonable the taxpayer’s actions were or how close a call the case was seems draconian to say the least, and in 1988 the government actually reversed course on this very issue and withdrew the original proposal for a GAAR-specific penalty. Indeed, nothing in the 1988 extrinsic aids references deterrence as an objective of GAAR. It would certainly be useful for the government to explain why the gross negligence penalties in s. 163(2) are inadequate for ensuring that taxpayers whose actions are clearly abusive are not sufficiently deterred.
As the government noted in the Discussion Paper, it has various options open to it for better articulating the legislative rationale of the provisions it enacts. More effort in this area would achieve vastly greater results without undermining the existing GAAR jurisprudence or increasing the risk of administrative over-reach by the CRA in applying GAAR where it ought not to (which creates costly and time-consuming controversies between taxpayers and the CRA). It is hoped that the government will concentrate on the real reasons it is not winning a greater percentage of the GAAR cases it chooses to litigate, and reconsider and refine the March 2023 Proposals accordingly. Interested parties should ensure that the Department of Finance receives their input by the May 31 deadline for making submissions, as it appears that no further consultations will occur thereafter before legislation is enacted.
For more information on the GAAR amendment submissions deadline, please reach out to one of the key contacts listed below.
1 Expressed by the SCC as “the rationale that underlies the words that may not be captured by the bare meaning of the words themselves”: Copthorne Holdings Ltd. v. Canada, 2011 SCC 63, at para. 70.