What is "treaty shopping"? It is an often used and disparaging phrase in international tax practice. But, in fact few really know what it means, contrasted with what for various reasons they want it to mean. Today, we have a much better idea of what "treaty shopping" is not. We are also reminded in that particular context but with far broader implications that tax treaties are self-interested bargains from each treaty partner's perspective, and that it is not within the remit of general anti-avoidance legislation to change or disavow the contractual terms or effects of treaties to which official exception is later taken.
The Supreme Court of Canada of Canada released its much-anticipated decision and reasons in The Queen v. Alta Energy Luxembourg S.A.R.L., 2021 SCC 49 (https://decisions.scc-csc.ca/scc-csc/scc-csc/en/item/19089/index.do)
The essential factual context is quite straightforward. Alta Energy Luxembourg S.A.R.L., a Luxembourg corporation (Alta), sold shares of a Canadian affiliate engaged in natural resource activities in Canada. It sold those shares not long after it acquired them in the course of a reorganization that entailed a taxable (by Canada) transfer by their former United States corporate owner to Alta. Alta had few commercial or other connections to Luxembourg, though both the taxpayer and the Canadian tax authorities agreed that by reason of having been incorporated in Luxembourg and having its “corporate seat” there it was resident in Luxembourg in the sense of "residence" contemplated by the Canada-Luxembourg Income Tax Convention (Convention). When it sold its shares of its Canadian subsidiary so acquired, not long after acquiring them, Alta relied on Article 13(4) of that Convention to deny Canada’s residual entitlement under the Convention to tax direct and indirect gains on dispositions of real property. The Tax Court of Canada decided, and it was as a result uncontested that the Canadian subsidiary whose shares were sold carried on its Canadian business in the natural resource properties the value of which was reflected in the gain realized on the sale of the shares. This satisfied an exception, in Article 13(4) of the Convention, to Canada’s entitlement otherwise to tax the gain based on the use of the real property essentially as business premises. The Federal Court of Appeal did not disturb the Tax Court’s finding that Alta was entitled to rely on the Convention in this way despite the Crown’s allegations of untoward and impugnable tax avoidance relying on the implications of the General Anti-Avoidance Rule (GAAR) in section 245 of the Income Tax Act (Canada) (Act) for how the Act as well as tax treaties should be interpreted and applied.
In the Supreme Court, the Crown did not contest Alta’s Luxembourg residence in the contemplation of the Convention or that its Canadian subsidiary carried on its business in the pertinent natural resource properties. Instead, relying on the GAAR, the Crown contended that the outcome of the transactions by which Alta came to own and sell the shares of the Canadian subsidiary was tax treaty abuse that ought to be impugned on the basis, as the Supreme Court framed the issue in paragraph 28 of the majority's reasons, that considering “the object, spirit, and purpose of the relevant provisions of the [Convention] … the avoidance transaction in this case …[resulted] in an abuse of those provisions…”. That was the issue faced by the Supreme Court in this case.
In a split six – three decision for the taxpayer, the Supreme Court disagreed with the Crown. The majority expressed themselves forcefully in careful, cogent, and complete legal analysis to deny GAAR as an instrument by which a treaty partner, by judging the laws of the other treaty partner wanting in some way according to how it saw things, could use purpose based GAAR arguments effectively to rewrite a treaty and the other treaty partner's laws. In effect, the majority recognized and made much of the fact and law that a tax treaty is a contract entered into by its parties to serve the interests of each, recognizing an intersection and a negotiated compromise of those interests. Consequently, the motive or purposes of the treaty partners were not to be second guessed or supplanted by a one (treaty partner) - sided anti - tax-avoidance based constitution or indeed reconstitution of transaction parties' (including treaty partners for the reconciliation of whose interests as countries tax treaties exist) supposed purposes in so acting, even in a situation like the present where the admitted tax avoidance objectives of the taxpayer were both clear and admitted.
This is perhaps one of the most important international tax cases decided in some time, anywhere by any adjudicators. It deserves and will get more considered comment in this Blog. But for the time being it may be helpful and certainly timely to identify some of its main elements that will be of interest to the international tax community and, no doubt, that will elicit much commentary shortly.
The majority engages in a careful analysis of the contractual nature of treaties in their public international law context, as they assess the interests pursued by both Canada and Luxembourg in having a tax treaty. The majority expressed this fundamental international tax truth as cogently and directly as imaginable in paragraphs 2, 4, and 9 of their reasons. They were reacting to their perception that the Crown now sought to disavow or change a treaty that had been properly concluded, with eyes wide open about the nature of Luxembourg's tax system, that country’s general characteristics pertinent to industry, trade, and commerce, and how taxpayers relied on the Convention, balanced by Canada's interest in attracting investment capital to the natural resource sector. The majority said:
[2] ... In the bilateral treaty context, there are two sovereign states whose intentions are relevant; a robust analysis must take both into consideration in order to give proper effect to the tax treaty as a carefully negotiated instrument. ...
[4] In my view, the Minister is asking this Court to use the GAAR to change the result, not by interpreting the provisions of the Treaty through a unified textual, contextual, and purposive analysis, but by fundamentally altering the criteria under which a person is entitled to the benefits of the Treaty, thus frustrating the certainty and predictability sought by the drafters. ...
[9] In raising the GAAR, Canada is now seeking to revisit its bargain in order to secure both foreign investments and tax revenues. But if the GAAR is to remain a robust tool, it cannot be used to judicially amend or renegotiate a treaty.
The majority recognizes and accepts the importance to Canada of encouraging real property and natural resource investment and finds that Canada is not entitled later to recast the treaty bargain in an image it did not have when it was entered into, according to considerations and perceptions of relevant law that were not in the minds of the treaty partners when they struck the bargain.
In taking the tack it did, provoked as it seemingly was by the Crown's attempt to use the GAAR to deny the Convention's application without having to engage with the difficult public international law and resulting reputational and implications associated with the breach or termination of an international agreement, the majority also may have fundamentally altered and limited the scope of the GAAR more generally. Also so affected, possibly, may be the anti-tax avoidance aspects of the Act’s transfer pricing provision in paragraphs 247(2)(b) and (d) of the Act, and as noted more below of the relatively new "principal purpose test" framed by the "Multilateral Instrument" to ground the selective denial of treaty benefits. Again, the majority's strong and categorical statement of their views in paragraph 96 of their reasons leaves little to the imagination:
[96] A final note on the Minister's implication that treaty shopping arrangements are inherently abusive. A broad assertion of "treaty shopping" does not conform to a proper GAAR analysis. In accordance with the separation of powers, developing tax policy is the task of the executive and legislative branches. Courts do not have the constitutional legitimacy and resources to be tax policy makers (Canada Trustco, at para. 41). It is for the executive and legislative branches to decide what is right and what is wrong, and then to translate these decisions into legislation that courts can apply. It bears repeating that the application of the GAAR must not be premised on "a value judgment of what is right or wrong [or] theories about what tax law ought to be or ought to do" (Copthorne, at para. 70). Taxpayers are "entitled to select courses of action or enter into transactions that will minimize their tax liability" (Copthorne, at para. 65). The courts' role is limited to determining whether a transaction abuses the object, spirit, and purpose of the specific provisions relied on by the taxpayer. It is not to rewrite tax statutes and tax treaties to prevent treaty shopping when these instruments do not clearly do so.
There is a careful consideration in the majority’s reasons of treaty residence and how it is and ought to be determined. Related to this, the majority touches on the possible significance of “beneficial ownership” as a potential, but in this case absent, limitation on the reach of treaty benefits including in respect of capital gains. In particular, assertions by the Crown that there needed to be meaningful degrees of economic, commercial, and other connections of a taxpayer with a jurisdiction to be considered resident there were not accepted. In this case, according to Luxembourg law, it was both acknowledged and admitted all around that Alta was as taxable as possible even though it was not subject to meaningful tax. This is another and an important tax treaty distinction frequently not well understood - between being liable versus subject to tax - that the majority addressed. Although not advanced as a “bumper sticker” mantra, the Court was astute to and remarked on notions of simplicity serviced by residence determinations, even of Canada, that look to where corporations are incorporated and do not impose external and undefined expectations of how or where their business activities are or ought to be conducted in particular as conditions to the allocation of taxing rights between treaty partners.
The majority also consider the relevance of model tax treaties and their evolved commentaries, notably of course the OECD Model Tax Convention and its Commentaries which, the majority carefully noted had evolved on matters of tax treaty – associated tax avoidance from when the Convention had been entered into. The issue of when and to what extent so-called subsequent commentaries are probative is important but rarely addressed in a judicial setting; the Canadian Prévost Car case is an important exception. In a world when treaty commentaries (including the OECD Transfer Pricing Guidelines in relation to Article 9 of the OECD Model Tax Convention) are experiencing dynamic and organic development, not only themselves but also in collateral proceedings associated with ongoing OECD-led international concerns about “base erosion and profit shifting” ("BEPS" and its "Pillars" progeny), this aspect of the majority’s reasons is notable generally. The majority do not dismiss outright the possible significance of later commentaries and other collateral guidance, but they are suitably cautious, again given the contractual nature of treaties and the importance of knowing and not embellishing what the treaty partners had and could only have had in mind when striking their bargain.
The majority’s reasons are “legal” in the most thorough and substantively respectful sense. They speak as refined and complete legal analysis. They are not political or economic, and do not admit the influence of untethered notions of “substance” on what the law, notably in a contractual setting, not only permits but requires to be decided. Although both the majority and minority reasons acknowledge the entitlement of taxpayers to engage in tax minimization planning – even that may be characterized as tax avoidance - the majority’s views are not an endorsement of the motor mechanics of whiteboarding (my turn of phrase, not the Court’s) to justify results simply because they have been bolted together in a particular way. The majority understand the importance of getting at the intrinsic expectations of legal constructions and, importantly, the legal bridge between two different legal and tax systems that is manifest in a public international law contract called a tax treaty. In this context, the majority also went out of its way to dispel any notion that somehow exogenous judgments about the morality of transactions and their tax outcomes are appropriate or relevant determinants of how to interpret and apply the Act and treaties, relying on earlier observations of the Supreme Court in the Copthorne Holdings case to which this Court refers.
The minority takes an entirely different tack in reacting to the taxpayer’s planning and the significance of the Convention in relation to it. Largely on the basis of their view of the absence of a meaningful presence of Alta in Luxembourg in relation to the denial of the Canadian tax liability by the Convention, they are concerned with what commonly in the international tax and GAAR contexts all too often would be referred to merely as “substance” and the entitlement to overcome or overlook the law based on contrary judgments about what the substance of events is or ought to be seen as. There is a prevailing tendency notably in relation to the OECD-led criticisms of “base erosion and profit shifting”, especially when “transfer pricing” is engaged, to consider the “accurate delineation” (from the 2017 OECD Transfer Pricing Guidelines) of the substance of otherwise effective transactions as the key and primary determination of how tax laws and what aspects of them should be applied in an international setting. The minority’s view is consistent with that approach, but the majority not only d0not embrace it but effectively characterize it critically to be the polar opposite of what they consider to be the proper approach to treaty interpretation and, it may be inferred, the interpretation and application of tax law more generally. Not unqualified, with a proper evidence based contextual examination of relevant circumstances still required. But not an untethered substance over form recharacterization of what actually and at law happened in favour of something else with some other consequences.
Another point of significance in this case for what is going on in the international tax conversation, is its possible effect on how the OECD BEPS - grounded “Multilateral Instrument” would succeed in superimposing on "Covered Tax Agreements “the requirement that the principal purpose of events in relation to tax treaties be determined in order to decide whether to respect the application of treaties as they otherwise would apply. The Alta case is, essentially, all about the purpose of tax treaties. The majority’s reasons, in subtle ways, reflect attention to tax treaty purpose in a way that seemingly could accommodate the implicit underpinnings of the PPT (“Principal Purpose Test”) in the Multilateral Instrument. But on closer examination, the legal analysis in this case and the qualified relevance accorded to successive OECD Model Tax Convention Commentaries dealing with tax avoidance, anti-avoidance notions embedded in treaties, and specific anti-avoidance rules, may be the source of legitimate and justifiable reservations about whether the PPT is as broad and elastic an anti-“treaty shopping” notion as many suppose or contend - or for that matter whether the PPT is anything other than what treaties imply without it.
In the days to come I anticipate there will be much said about this case, including here. For the moment, the significance of this case deserves to be noticed on the same day the Supreme Court has spoken, even though inevitably there is more to come and more should come.
Scott Wilkie
Distinguished Professor of Practice, Osgoode Hall Law School, York University