At the current stage, it is difficult to predict a potential application of the principal purposes test (PPT) and its outcome since it has not entered into force in the 72 Signatories of the Multilateral Convention to Implement Tax Treaty Related Measures to Prevent Base Erosion and Profit Shifting (MLI). However, assuming that this happens in the near future, an attempt can be made to disperse some of the confusion that arose among practitioners and scholars in respect of the potential application of the PPT – the procedural issues regarding burden of proof. It will be argued that the criticism towards an alleged shift of burden of proof from the tax authorities to taxpayers under the PPT is not well justified and may be misleading for readers. Also, contrary to the views of many authors, there seem to be no real difference between burden of proof under the PPT and the “guiding principle” of para. 61 of the Commentary on Art. 1 of the 2017 OECD Model Convention (before: para. 9.5 of the Commentary on Art. 1 of the OECD Model Conventions as released in 2014, 2010, and 2003). Finally, the author postulates de lege ferenda will be delivered with the aim to balance the interest of the tax authorities with that of the taxpayers.
The exacerbated criticism of practitioners and scholars regarding the PPT’s burden of proof
The procedural issues regarding burden of proof under the PPT sparked criticism from many practitioners and scholars. Since their criticism may be confusing and misleading for readers, it is worth briefly discussing it.
Lang seems to be the first author to express his disapprobation with the PPT’s approach to the burden of proof:
If, as part of its official duty of investigation, the tax authority must furnish proof that one of the main objectives of the taxpayer was to obtain the benefit, it is already fighting a losing battle. Alternatively, the taxpayer has no chance of fending off the accusation of abuse if it is up to him to furnish evidence that benefiting from one or several treaty provisions was not one of his primary motives. [(…)] The rule [the PPT] thus attempts to establish a balance between the interests of the authority and those of the taxpayer. The bias in favor of the tax authority, however, is fairly obvious, a fact that was also critically commented in several statements submitted to the OECD. In practice, furnishing evidence of the motives will therefore not be relevant, but tax authorities will be tempted to presume intention simply because of the presence of a benefit. For this reason alone, the subjective criterion runs the risk of not gaining any significance in itself. [Italics added]
De Broe and Luts seem to be largely influenced by Lang’s position:
The PPT also inconspicuously alters the division of the burden of proof regarding treaty abuse. Under the previous ‘guiding principle’, it was common ground that the tax authorities of the State desiring to refuse treaty benefits carry the burden of proving that both the subjective and objective element of abuse are fulfilled. [(…)] Under the proposed PPT, the tax authorities are still obliged to ‘prove’ the presence of the subjective element, but only this. Moreover, the threshold is set extremely low: it is not required for them to ‘establish’ (beyond reasonable doubt) that obtaining a treaty benefit was one of the principal motives. Rather, it suffices that ‘it is reasonable to conclude’ that such motives were present. Although ‘an objective analysis of the relevant facts and circumstances’ is still required, an author has already concluded that ‘the subjective criterion runs the risk of not gaining any significance in itself’. [Italics as in the original version]
The essence of these findings has been repeated by many other authors, such as Pinetz, Koriak, Bergedahl, and Danon.
This basically creates an impression that the tax literature on the PPT, apart from the views of Taboada and Weber in respect of the standard of reasonability (and I agree with them), is currently convinced that the burden of proof under the PPT falls only at the beginning with the tax authorities. This regards the first element of the PPT, i.e. reasonably concluding that one of the principal purposes of a taxpayer’s arrangement or transaction was to obtain a treaty benefit. Then the burden of proof is effectively shifted onto taxpayers under the second element of the PPT, i.e. establishing by a taxpayer that the obtainment of the treaty benefit was in accordance with the purpose of the relevant provisions of the tax treaty. This could undo the balance between tax authorities and taxpayers in favour of the former, because it is much easier to prove the first than the second element of the PPT. The authors’ findings, however, are confusing and not very cogent insofar as they neither explain what they mean by “burden of proof” under the PPT nor do they acknowledge different approaches to procedural aspects of burden of proof in tax cases in various countries. Nor do they distinguish between the nature of the first and of the second element of the PPT.
A closer look at “burden of proof” in the light of the PPT’s burden of proof: Not as black as it is painted
The literature on burden of proof in tax law generally agrees that a burden (onus) is a question of law while a proof relates to plain facts.
In respect of the PPT, one may say that the questions of proof are related to the relevant facts and circumstances on which the consideration of one of the principal purposes of the taxpayer’s arrangement or transaction shall be based (the first element of the PPT). On this view, the burden of proof is neither on the tax authorities nor the taxpayer, but split between them: the tax authorities must make the first move by gathering and demonstrating evidence that one of the principal purposes of the taxpayer’s arrangement or transaction was to obtain a treaty benefit, and then the taxpayer follows through by gathering and demonstrating evidence to the opposite (e.g. the existence of economic substance and business purposes other than tax avoidance). So if the tax authorities have no indication that one of the principal purposes was to obtain treaty benefits, there will be no reason to apply the PPT and taxpayers will not need to defend themselves. In that sense, the burden of proof rests essentially with the tax authorities.
It is noteworthy that the “split” converges with national rules on burden of proof under anti-abuse provisions: in a first step, the tax authorities have to prove that the legal arrangement in question is abusive according to national rules (e.g. a GAAR); it is then within the sphere of the taxpayers to provide counterevidence. This is also in line with the general rule on the burden of proof among the reported countries of the EALTP (European Association of Tax Law Professors) annual meeting held in Uppsala from 2–3 June 2011 on “The Burden of Proof in Tax Matters”, apart from Austria, since Austrian law generally puts the burden of proof on the tax administration. For this reason, perhaps, Lang, as an Austrian scholar, being influenced by domestic procedural rules on burden of proof, criticized the drafters of the PPT for switching the burden of proof from the tax authorities to the taxpayers.
As for the questions of burden (questions of law), they appear to be pertinent to the consideration of the compatibility of the taxpayer’s arrangement or transaction with the purpose of relevant treaty provisions (the second part of the PPT which includes negative condition of the PPT). This functions as a defensive rule in a sense that when it is reasonably concluded that one of the principal purposes of the taxpayer’s arrangement or transaction was to obtain a treaty benefit, the taxpayer has the burden of argument, which is a question of law, that the obtaining this benefit is in accordance with the object and purpose of the relevant treaty provisions.
In common law countries, tax law stipulates that questions of burden (the statutory onus) lie initially with taxpayers. Hence, in these countries, the question of burden under the PPT will not be considered as a shift of the statutory burden to a taxpayer. Neither will it be seen as a shift in civil law countries where anti-abuse rules are applied, since they usually allow the burden to be switched from the tax authorities to taxpayers in cases targeted by these rules. Nor will it be a significant issue in civil law countries without anti-abuse rules insofar as their procedural laws tend to split the statutory burden rather than putting it on the tax authorities (except Austria).
Accordingly, criticism of the authors mentioned above does not appear to have much legal basis. It rather seems to follow from their overall disapprobation towards a tax treaty GAAR construed as the PPT.
The PPT’s burden of proof vs the guiding principle’s burden of proof: Is there really a difference?
Still, their criticism may be seen as valid to the extent that it challenges the OECD’s statement about the PPT’s mirroring of a guiding principle. In addition to obvious linguistic differences, the guiding principle requires the tax authorities to prove whether the taxpayer’s transaction or arrangement mainly aims to obtain treaty benefits and determine that this is contrary to the purpose of the relevant treaty provision. Only if both requirements are met by the tax authorities, may they proceed to deny a benefit under the tax treaty provision(s), while a taxpayer may try to stop this denial by proving and determining otherwise. Under the PPT, in contrast, it is enough that the tax authorities prove that one of the principal purposes of the taxpayer’s transaction or arrangement was to obtain treaty benefits to proceed to deny the benefits. The way to deny treaty benefits appears to be shorter and easier under the PPT in comparison to a guiding principle. But is this really so the case?
Devolving the first element on the tax authorities and the second on the taxpayer does not change the finding that both elements of the PPT matter equally to denials of treaty benefits. As the examples regarding the application of the PPT in the paras. 182 and 187 of the Commentary on the Article 29(9) of the 2017 OECD Model Convention reveal, a reasonable conclusion that one of the principal purposes of the taxpayer’s transaction or arrangement was to obtain treaty benefits is coincidental with establishing that it was contrary to the purpose of relevant treaty provisions. That is to say, the first and second elements of the PPT are strictly interrelated. It is therefore very unlikely that the tax authorities will try to deny treaty benefits by concluding that the requirement for doing so is met under the first element while being assured that the second element allows taxpayers to neutralize this denial. Indeed, the tax authorities will take into account the second element of the PPT at the concluding stage of their first element consideration, even though the wording of the PPT does not require them to do so. The need to consider the purpose of tax treaties according to the rule of interpretation under Article 31(1) of the Vienna Convention on the Law of Treaties (VCLT) supports this observation too.
On this view, the authors’ claims that only the second part of the PPT will actually matter do not gain much of legal and pragmatic importance. Moreover, their underlying arguments are not very cogent either. Namely, they say that while taxpayers typically think of obtaining tax treaty benefits, tax authorities will have few difficulties in assuming that one of the principal purposes of the taxpayer’s transactions or arrangements is precisely to obtain treaty benefits. It is obvious that reasonable taxpayers who are engaged in cross-border businesses and investments are aware that entering into transactions or arrangements to which a tax treaty applies secures a more favourable tax position. And clearly, they use that knowledge to weigh up the most favourable route with the use of all legally permissible and financially rational options, including the award of treaty benefits. It is also obvious, however, that a considerable amount of tax treaty benefitted transactions happen as a matter of routine. This regards, for example, cross-border dividend payments by large listed public corporations in respect of institutional share portfolios of pension funds, exchange traded funds, and alike. Although all these payments will benefit from the tax treaty’s equivalents of Article 10(2)a of the 2017 OECD Model Convention (reduced withholding tax on dividend payments), can one really conclude, as the authors do, that “one of the principal purposes” test can be simply assumed by tax authorities against such taxpayers for these types of payment, and so not really matter? In addition to the above findings, Example E in the para. 182 of the Commentary on the Article 29(9) of the 2017 OECD Model Convention undermines the creditability of such a conclusion.
One may make a valid point de lege ferenda, though, that in order to approximate the way the PPT functions like a guiding principle, the word “unless” could be replaced with “and” under the PPT. This would not only switch the entire burden of proof initially to the tax authorities, but also explicitly require them to take into account the purpose of relevant treaty provisions from the very beginning. Since there is actually no guideline in the Commentary on the Article 29(9) of the 2017 OECD Model Convention on how to determine such a purpose, requiring tax authorities to determine it first may help taxpayers while counter-arguing on that matter, i.e. they may use the tax authorities’ approach as a hint in determining the purpose of relevant treaty provisions. This would also help balance the interest of the tax authorities with that of the taxpayers.
 See M. Lang, BEPS Action 6: Introducing an Antiabuse Rule in Tax Treaties, Tax Notes International, May 19, pp. 658–659 (2014).
 See De Broe & Luts, supra n. 4, pp. 132.
 See E. Pinetz, Final Report on Action 6 of the OECD/G20 Base Erosion and Profit Shifting Initiative: Prevention of Treaty Abuse, 70 Bulletin for International Taxation 1/2, pp. 115–120 (2016); O. Koriak, The Principal Purpose Test under BEPS Action 6: Is the OECD Proposal Compliant with EU Law?, 56 European Taxation 12, p. 557, (2016); Ch. Bergedahl, Anti-Abuse Measures in Tax Treaties Following the OECD Multilateral Instrument – Part 1, 72 Bulletin for International Taxation 2, Published online: 9 January 2018, sec. 22.214.171.124; R. Danon, Treaty Abuse in the Post-BEPS World: Analysis of the Policy Shift and Impact of the Principal Purpose Test for MNE Groups, 72 Bulletin for International Taxation 1, 2018, Published online: 28 December 2017, secs. 4.1, 4.4.1, 4.4.2, and 4.5.
 See C. Palao Taboada, OECD Base Erosion and Profit Shifting Action 6: The General Anti-Abuse Rule, 69 Bulletin for International Taxation 10, p. 604–605 (2015); D. Weber, The Reasonableness Test of the Principal Purpose Test Rule in OECD BEPS Action 6 (Tax Treaty Abuse) versus the EU Principle of Legal Certainty and the EU Abuse of Law Case Law, 10 Erasmus Law Review 48, p. 51 (2017).
 See G. Marino, The Burden of Proof in Cross Border Situations (International Tax Law), in The Burden of Proof in Tax Law, G. Meussen (ed.), the book based on the proceedings of the annual meeting of the European Association of Tax Law Professors (EATLP) held in Uppsala from 2-3 June 2011, section 4.1, (2013).
 See B. Leidhammer, National concepts, in Meussen supra n. 4, section 1.1 and
J. Heinrich, Austria, in Meussen supra n. 77, section 1.
 See, for example, in US see H. Ordower, United States of America, in Meussen supra n. 77, section IV. A.1, and in the UK see L. Fichardt & J. Seddon, Burden of proof, in Tax litigation in the UK (England and Wales): overview, available online at: https://uk.practicallaw.thomsonreuters.com/1-623-2951?transitionType=Default&contextData=(sc.Default) (accessed 18/1/2018).
 See K-D. Drüen & D. Drissen, Burden of Proof and Anti-Abuse Provisions, in Meussen supra n. 4, section 2.1.1.
 See Leidhammer, supra n. 5, section 1.1.
 Cf. comments of Taboada on Lang’s criticism regarding the PPT in Taboada, supra n.4, pp. 606–608 (2015).
 See the Examples A–J (general treaty abuse examples) and A–F (conduit treaty abuse examples).
 See Lang supra n. 1, p. 658–659; De Broe & Luts, supra n. 2, pp. 132; Pinetz, supra n. 3, p. 116. Cf. P. Baker & T. Liao, Improper Use Improper Use of Tax Treaties: The New Commentary on Article 1 and the Amended Article 13(5), 66 Bulletin for International Taxation 11, p. 600 (2012),; G. S. Cooper, Tax Treaty Policy of Developing Countries post-BEPS, 4 School of Accountancy Research Paper Series, p. 19 (Paper No: 2016-S-45, 2016).