1. The IIR and the top-down approach
It has become commonplace that the current OECD and Inclusive Framework (IF) proposals to address the tax challenges of the digital economy will represent a fundamental overhaul of the international tax system as we know it. As time passes, the international community has time to digest these possible changes and assess just how fundamental they will be. Systematization efforts, therefore, will be key to ensure a uniform understanding of the rules across the globe. In particular, they will have to deal with contradictions related to the fact that the practical perspective of multinational enterprise groups (MNEs) as cohesive units clashes with the fundamental reliance of the Global Anti-Base Erosion (GloBE) Rules on the separate entity approach.
At a high level, the policy objective of Pillar 2 has been clear from the outset: to “strengthen the ability of jurisdictions to tax profits where the other jurisdiction with taxing rights applies a low effective rate of tax to those profits”.[1] In other words, the goal is to set a floor to tax competition – defined at a 15% Effective Tax Rate (ETR).[2]
To achieve this objective, the OECD/IF proposed the GloBE Model Rules, which create a Top-up Tax liability connected to the low-taxed income of a Constituent Entity. In this way, jurisdictions that could be tempted to engage in tax competition would be discouraged to do so, in the knowledge that the Constituent Entity would not benefit from any incentive – since any taxation lower than 15% ETR would be captured in another jurisdiction.[3] Therefore, the Top-up Tax encapsulates a powerful tool for the standard setters to nudge the rest of the world towards higher ETRs.
Operationally, the GloBE Model Rules include two interlocking rules: the Income Inclusion Rule (IIR) and the Undertaxed Payments Rule (UTPR). While the UTPR plays quite an important role, our discussion focuses on the IIR Top-up Tax.
More specifically, the IIR applies on a top-down basis.[4] According to article 2 of the Model Rules, priority is given to the application of the Top-up Tax in the Ultimate Parent Entity (UPE). Only if the UPE is not required to apply the IIR, then an intermediate parent Entity would be required to do so (with priority to intermediate parents higher up in the corporate chain).
As for their implementation, the GloBE rules are intended to have the status of a “common approach”. This means that IF members: (i) are not required to adopt the rules, but if they choose to do so, the enacted rules must be consistent with model rules and guidance agreed to by the IF; and (ii) accept the application of the GloBE rules by other IF members, including agreement as to rule order and safe harbors.[5] This also means that the implementation of the GloBE rules is not intended to require any changes to the bilateral treaties and the rules could be applied merely by means of amendments to domestic legislation.[6] It is therefore assumed by the IF that the GloBE rules are compatible with the existing double tax convention (DTC) framework – an assumption that, we will argue, may be overly optimistic.
2. What is the nature of the IIR?
From its name, it seems reasonable to derive that the IIR operates as a mechanism to ensure the taxation of income. But whose income is being taxed, and how come would it be taxed in the parent Entity jurisdiction? The IF work on the topic allows for two possible answers to this question.
2.1 The Blueprint-Answer: “inclusion” of low-taxed income as income of a parent
Following the Pillar Two Blueprint, the IIR would require the inclusion of foreign-source income of the foreign Entities of the MNE group as income of the parent Entity[7] – therefore being called an “Income Inclusion Rule”. The IIR would effectively operate by requiring a parent Entity (in most cases, the UPE) “to bring into account its share of the income of each Constituent Entity located in a low-tax jurisdiction”, taxing the income up to the Minimum Rate.[8]
The Blueprint-Answer is openly inspired by the mechanism of Controlled Foreign Corporation (CFC) rules.[9] This mechanism is generally triggered in abusive situations, in which the income of the foreign subsidiary is presumably accrued with aid of abuse. In such circumstances, CFC rules would enable the income to be included and taxed as income of the parent Entity – where, in the absence of abuse, it would have presumably been registered.
The IIR would operate in much the same way, except that no abuse is assumed.[10] Income legitimately attributable to a foreign subsidiary, but taxed at a low rate, would be included at the level of the parent Entity.
This background raises a fundamental question about the nature of the Blueprint-IIR: could it be deemed equivalent to a tax on the income of the parent Entity?
CFC rules fit the notion of an income tax because of its anti-abuse character. Rightly or not, the underlying assumption is that the income of the foreign subsidiary should actually be considered to be income of the parent entity. Incidentally, this is the justification given by the OECD Commentary, since 2003, to support the alignment of CFC rules with article 7 of tax treaties: these rules tax the income of the parent, not of the subsidiary.[11]
An alternative approach could seek inspiration in the Brazilian full inclusion system: regardless of any abuse, the parent entity would register in profit and loss (P&L) the equity variation derived from the investment in the foreign subsidiary. This approach would not deny that the income is correctly attributable to the foreign subsidiary, but it would also allow the inclusion of the same income at the level of the parent entity. The result would, essentially, lead to (economic) double taxation – except that relief would be provided up to the level of the minimum tax.
As it happens, however, the IIR gives priority to the allocation of the Top-up Tax to the UPE, which has a much further connection to the low-taxed income than the direct parent Entity. The difference is quite significant: if the direct parent Entity, in addition to the profits derived from equity variations from the investment in the subsidiary, also incurs in losses, then no income would be passed further along to the UPE. In other words, the reflection of the subsidiary’s profits in the P&L of its direct parent does not necessarily impact the P&L of the indirect parent (assuming separate entity accounting, which is the starting point of the IIR). In this way, the IIR at UPE level could be levied over income that would not necessarily become income of the UPE.
The Blueprint-Answer presents, therefore, significant challenges, be it because of the lack of anti-abuse character, be it because the top-down approach is inconsistent with the computation of the UPE’s income. Besides, the reasoning would also fail to explain the nature of the UTPR, which should work as a “backstop” for the IIR – therefore grounded on a similar reasoning.
2.2 The Model Rules-Answer: the “inclusion” of low-taxed income as income of the Low-Taxed Constituent Entity
Much has changed in Pillar 2 between the 2020 Blueprint and the 2021 GloBE Model Rules – so much so that the answer to the fundamental question of who is the relevant taxpayer may have changed.
A close examination of the charging provision in the Model Rules reveals that the “income” being taxed is that of the LTCE, and that no “inclusion” in the parent Entity’s tax base is provided (so that IIR is, in the Model Rules, a “misnomer”[12]). According to Article 2.1, the UPE “shall pay a tax in an amount equal to its Allocable Share of the Top-Up Tax of that Low-Taxed Constituent Entity [LTCE] for the Fiscal Year”. The Top-Up Tax, therefore, is calculated by reference to the income of the LTCE (with certain flexibility for jurisdictional blending). The reference to an Allocable Share is intended to recognize that the liability at UPE level should be proportional to the Ownership Interests held in the LTCE. Moreover, the wording of the rule indicates that the UPE shall “pay” the tax – it does not suggest that “income” is as such attributable to the UPE.
This construction seems at odds with the Blueprint-Answer. Rather, the wording of the Model Rules indicates that the LTCE itself would be the taxpayer of the IIR (so understood as the Entity that gives cause to the taxable event). The parent Entity would act merely as a third party collecting agent (it does not itself incur in the taxable event, but the liability is assigned to the UPE or the intermediate parent, given its mutual relationship with the LTCE and the IIR jurisdiction).
This reasoning entails, however, recognizing that the country of residence of the parent Entity has jurisdiction to tax income with no objective or personal connection to its territory. The income is accrued in a different jurisdiction and to a different taxpayer, but the parent jurisdiction would nonetheless find itself in a position to tax it. Such an approach would not only conflict with the logic of tax treaties, but potentially also with the current notion of jurisdiction in international law.
The first problem is that, as acknowledged under tax treaties, the non-residents can only be taxed in case a nexus with the source jurisdiction is recognized – either a Permanent Establishment or other specified types of income that are earned in the source jurisdiction. The IIR under the Model Rules-Answer breaks with this logic, by allowing taxation of non-residents without the mentioned nexus with the IIR jurisdiction. The incompatibility of the approach with the current DTC framework seems inevitable, and it is hard to see how the IF expects to apply it without a fundamental overhaul of existing bilateral treaties.
The second problem is that, from an international law perspective, the very notion of jurisdiction has to be revisited in the Model Rules-Answer. In his lectures, Prof. Luís Schoueri makes a point by asking students whether Brazil can levy a property tax on the White House. The negative answer is not attributable to the absurdity of the idea, but to the notion of jurisdiction: Brazil lacks jurisdiction to levy a tax on property situated outside its territory (and we do not need a tax treaty stating that).[13]
Likewise, if an Entity has no objective or personal connection with a certain jurisdiction, there seemed to be no reason why the profits of the Entity would be subject to taxation in such jurisdiction (and, until the discussions under Pillar Two, we did not seem to need a tax treaty stating that either). It is true that the CFC reasoning has, to a certain extent, eroded this understanding, resorting to fictions and deeming provisions to tax the income of the subsidiary as income of the parent, for anti-abuse reasons. However, the IIR represents a step further: a top-up tax liability may apply even if no abuse can be identified and no constructive income attribution to a resident entity takes place. As there is no attribution of income to a resident entity, the IIR implies the creation of a jurisdiction to tax an LTCE, regardless of the existence of objective or personal connection to the territory. This would be a fundamental change in the notion of jurisdiction as we know it. Whether such an approach is in violation to any (customary) principle of international law will certainly become a topic of interest in academic fora in the forthcoming years.
3. Who is the “taxpayer” in the GloBE Model Rules?
Despite the rhetoric grounded on MNEs as cohesive units, the Model Rules are strongly reliant on the separate entity approach. As a consequence, the identification of the taxpayer still requires the reference to legal entities and the reference to an MNE Group does not suffice. The evolution of Pillar 2 suggests two possible answers to the question we pose.
Under the Blueprint-Answer, the taxpayer would be a parent Entity. The income of the LTCE would be “included” as income of the parent Entity and taxed in the parent’s jurisdiction. This answer, while inspired on CFC rules, would have raised significant challenges derived from the lack of (anti-abuse) reason for attributing the LTCE’s income to another Group Entity.
Under the Model Rules-Answer, the taxpayer would be the LTCE (and the parent Entity would be merely a collecting agent). In this case, the parent’s jurisdiction has a right to tax income of the LTCE, despite the inexistence of objective or personal connection of the LTCE with the jurisdiction’s territory. The wording of the Model Rules indicates that the Top-up Tax is calculated with reference to the LTCE, and merely collected by a parent Entity (IIR) or another Group Entity (UTPR). This answer presents many challenges related to its compatibility with the traditional understanding of tax jurisdiction and with the intent to make it compatible with DTCs.
As posed, the issue brings important questions which will have to be addressed in legal scholarship – the most immediate of which would include:
- How will the IIR interact with constitutional systems and could its misalignment with the notion of income of the UPE present additional implementation challenges?
- How will the IIR interact with tax treaties and can the saving clause protect the IIR from article 7? Skepticism in this regard has already been expressed by other authors,[14] including in this blog.[15]
- What are the limits for the jurisdiction to tax corporate profits under international law, if any?
The answer will have to consider the UTPR as well. Despite the scope limitation of the present note, finding a consistent answer will require considering the Model Rules as a whole, as the IIR and the UTPR are two interlocking rules, and, as such, should be grounded on a common logic. The UTPR brings additional questions, which the present note does not consider.
However, such a principled approach and systematic view of the GloBE Rules do not seem to have been priorities in the (political) development of the rules. Curiously enough, the Model Rules contain two usages of the term “taxpayer” when referring to the deferred tax asset, but it remains unclear to whom the term refers.[16] The GloBE Commentary does not present any particular concern with the content of the term either, and the question we pose cannot be answered with reference to the miscellaneous usage found in the Commentary. The difficulty in answering even a basic question such as “who is the taxpayer” is a reason for significant concern. Further refinement of the GloBE principles will be key to ensure its smooth implementation by states.
[1] OECD (2019), Addressing the Tax Challenges of the Digitalisation of the Economy – Policy Note (As Approved by the Inclusive Framework on BEPS on 23 January 2019), p. 2..
[2] OECD (2021), Statement on a Two-Pillar Solution to Address the Tax Challenges Arising from the Digitalisation of the Economy, p. 4.
[3] For the purpose of this blog post, without prejudice to our reasoning, we leave aside some elements that would require further refinement, such as jurisdictional blending, the carve-out and the temporary adjustments.
[4] Articles 2.1.4 and 2.1.5 provide for special rules in the case of Partially-Owned Parent Entities (POPE), which are not addressed in the present note.
[5] OECD (2021), Statement on a Two-Pillar Solution to Address the Tax Challenges Arising from the Digitalisation of the Economy, p. 3.
[6] OECD (2020), Tax Challenges Arising from Digitalisation – Report on Pillar Two Blueprint, , para 21.
[7] OECD (2020), Tax Challenges Arising from Digitalisation – Report on Pillar Two Blueprint, para. 410.
[8] OECD (2020), Tax Challenges Arising from Digitalisation – Report on Pillar Two Blueprint, para. 410.
[9] See OECD (2020), Tax Challenges Arising from Digitalisation – Report on Pillar Two Blueprint, para. 411. For further analysis, see Johanna Hey, The 2020 Pillar Two Blueprint: what can the GloBE Income Inclusion Rule do that CFC legislation can’t do?, Intertax, vol. 49, issue 1, 2021, pp. 9-13.
[10] The issue has been discussed in more detail in Ana Paula Dourado, The Pillar Two Top-up Taxes: interplay, characterization and tax treaties, Intertax, vol. 50, Issue 5, 2022, pp. 388-395.
[11] OECD (2017), Commentary on article 7 of the OECD Model Tax Convention (Condensed Version), para. 14.
[12] Brian J. Arnold. The ordering of residence and source country taxes and the OECD Pillar Two Global Minimum Tax, Bulletin for International Taxation, vol. 76, No. 5, 2022, pp. 220.
[13] Luís Eduardo Schoueri, Direito Tributário, 11 ed., São Paulo, Saraiva, 2021, p. 111.
[14] Luis Eduardo Schoueri, Some Considerations on the Limitation of Substance-Based Carve-Out in the Income Inclusion Rule of Pillar Two, Bulletin for International Taxation, December 2021, pp. 543-548; , Vikram Chand, Alessandro Turina and Kinga Romanovska, Tax Treaty Obstacles in Implementing the Pillar Two Global Minimum Tax Rules and a Possible Solution for Eliminating the Various Challenges, World Tax Journal, 2022 (Volume 14), No. 1.
[15] Maarten de Wilde, Why Pillar Two Top-Up Taxation Requires Tax Treaty Modification, Kluwer International Tax Blog, January 12, 2022, available at https://kluwertaxblog.com/2022/01/12/why-pillar-two-top-up-taxation-requires-tax-treaty-modification/ accessed on 12/06/2022.
[16] See, GloBE Model Rules, Art. 4.4.3. and Art. 9.1.1.
________________________
To make sure you do not miss out on regular updates from the Kluwer International Tax Blog, please subscribe here.
