Stefanie Geringer, Researcher at the University of Vienna and senior associate at BDO Austria
The Court of Justice of the European Union (CJEU) has developed a new argumentative pattern attributing a dynamic relevance to the commentary on the model tax convention of the Organisation for Economic Co-operation and Development (OECD) for purposes of OECD-induced concepts in European Union (EU) secondary law. In light of the recent developments on a ‘global tax reform’ negotiated in the OECD Inclusive Framework, this case law should raise awareness (not only) amongst tax scholars in the EU and beyond.
1. The bigger picture: Controversies about the dynamic relevance of the commentary in the context of bilateral tax treaties
The argument on the use of a dynamic understanding of the commentary on the OECD’s model tax convention (hereafter convention) originates from international tax law.
The United Kingdom (UK) as well as EU Member States usually rely on the convention as a negotiation base for their bilateral tax treaties.[i] These treaties are traditionally aimed at defining to whiat extent each contracting partner shall be eligible to exercise their taxing rights. Like any other provision, these allocation rules raise interpretation questions. Many of these are addressed in the commentary on the convention. Both the convention and (particularly) the commentary are updated on a regular basis.
Probably as old as the convention itself is the discussion about which and how much value should be attached to which version of the commentary. It is broadly acknowledged that the commentary relevant at the time of signing a bilateral tax treaty can prove valuable for interpretation purposes as far as treaties follow the OECD model.[ii] More controversial, however, is the question of whether later versions of the commentary should be considered in the interpretive procedure. These discussions are nurtured by the fact that this issue is usually not dealt with in the treaties.[iii]
There are several arguments speaking for or against a dynamic interpretation. Without claiming to be exhaustive,[iv]the following arguments have been put forward to defend the exclusive use of the commentary existent at the time of signature:
- It serves legal certainty;
- complies with constitutional law (particularly the principle of separation of powers);
- ensures uniform application in the contracting states irrespective of changes to the commentary; and
- can be reasonably held to reflect the contracting parties’ intentions.
There are also reasonable arguments that can make the case for using commentaries published after the time of signature of the treaty in question:
- It allows for considering changes in the political, economic, legal and societal environments as well as technological advances;
- is able to meet legitimate expectations of taxpayers; and
- does justice to the principle of equality because OECD model-inspired treaty provisions are uniformly applied no matter when the relevant treaty was signed.
The merits of both sides of the argument are reflected in the highly divergent case laws in the UK, the EU Member States and beyond.[v]
2. Importing international tax issues into the EU legal sphere: the CJEU’s decisions in Berlioz, N Luxembourg 1 and Others and État luxembourgeois
Several concepts in the convention have been ‘borrowed’ for purposes of EU directives. Among them are the foreseeable relevance criterion in the Directive on Administrative Cooperation (DAC) and the beneficial ownership requirement in the Interest-Royalties Directive (IRD).[vi] The former was at the centre of the CJEU’s decisions in Berlioz (C-682/15) and État luxembourgeois (C-437/19), the latter in N Luxembourg 1 and Others (Joined Cases C-115/16, C-118/16, C-119/16 and C-299/16).[vii]
One of the preliminary questions in Berlioz addressed the issue whether the requested information must be foreseeably relevant to the administration and enforcement of the requesting state’s domestic tax laws to legitimize requests under the DAC regime. Beyond simply confirming this requirement, the CJEU further discussed the essence of the foreseeable relevance criterion. It derived its main findings from an autonomous interpretation in light of Recital 9 of the DAC. These were supported by references to related statements in the 2012 commentary. This is remarkable considering that the DAC was adopted in 2011, at a time when EU lawmakers could have naturally not yet been familiar with the 2012 commentary version. It appears even more remarkable against the background that these statements on the foreseeable relevance criterion had already been included in the 2005 version in an identical manner. The application of a dynamic understanding hence did not add any value, but unnecessarily raised similar issues as in the treaty context discussed above.[viii]
A similar argumentative pattern can be identified in N Luxembourg 1 and Others in the context of the beneficial ownership requirement of the Interest-Royalties Directive (IRD). First, the Court drew from a comparative analysis of the IRD’s 23 language versions that the beneficial ownership requirement needs to be understood in an economic sense. To bolster its interpretive result from an autonomous interpretation, the CJEU deemed the 1996 convention and commentary as well as successive amendments thereof relevant when interpreting the IRD. It then referred to statements in the 2003 commentary that proved particularly useful for its argument. Nevertheless, the use of a dynamic understanding in N Luxembourg 1 and Others would have been equally avoidable: Similar statements were already made in earlier OECD work[ix] that was indeed expressly mentioned in the commentary. At first glance, the Court’s relaxed approach towards potential deficiencies in democratic legitimacy related to the use of a dynamic interpretation are moreover striking.[x] This attitude might however stem from the fact that the CJEU uses references to the commentary merely to confirm interpretive results from an autonomous interpretation of EU law.
The État luxembourgeois case centered on the question of how precisely individual taxpayers ought to be described in an information request under the DAC regime (in light of the prohibition of ‘fishing expeditions’[xi]). In similar fashion as in the earlier cases, the CJEU built an argument on the basis of an autonomous interpretation of EU law. This interpretation in light of the wording, the legal context and the objectives pursued elucidated that a request related to a targeted investigation into a limited group of persons could be deemed sufficient. The Court then found confirmation of his argument in the 2012 commentary where the issue of group requests was explicitly addressed for the first time. Arguably, a doctrinally more convincing argument would have yet pointed to the character of these statements as mere clarifications of the foreseeable relevance criterion established in the 2005 convention.
A common thread in all three decisions is the absence of an explanation for the fundamental legitimacy of the use of a dynamic understanding of the commentary. This is remarkable considering that such an approach is already highly disputed at the international tax level (hence, the origin of the convention and its commentary).[xii]
3. Why should we care about the CJEU and national courts applying a dynamic interpretation of the commentary?
By using a dynamic understanding of the commentary in its recent case law on OECD-inspired secondary law provisions,[xiii] the CJEU imported the problems related to this approach into the EU legal sphere. The same effect is realized by national courts referring to later versions of the commentary in the interpretation of bilateral tax treaties or related domestic laws (eg transfer pricing rules). The issues arising from the CJEU’s stance are hence not only relevant for the EU Member States, but to all jurisdictions including the UK using OECD-inspired concepts in their national and treaty laws.
In the cases decided by the CJEU, it could be argued that all arguments were originally built on an autonomous interpretation of EU law so that presumably ‘no damage was done.’ Vice versa, the question should be raised: What’s then the value added from references to OECD documents issued after the adoption of the relevant (EU) legal sources?
Cutting the argument short by concluding that it does not do any harm misses another crucial point: the issue whether these OECD documents can be legitimately considered in the interpretive process in the first place. As long as the OECD’s decision-making process exclusively involves representatives from national domestic tax authorities that are not subject to any legislative control, direct references to later OECD documents might be problematic at least to some extent from an (EU) constitutional perspective. This issue is further exacerbated in the EU context, as 5 of the 27 EU Member States are not members of the OECD[xiv] and hence cannot influence the negotiation results materializing in the convention and its commentary. Against this background, each court referring to OECD work in its decisions – be it a national court or the CJEU – should make a solid argument for the legitimacy of its approach when using documents issued after the adoption of the relevant legal text. The lack of any such reasoning is what could arguably be most fiercely criticized about the CJEU’s new argumentative pattern in the context of interpreting OECD-inspired concepts of EU secondary law.
Considering that 137 tax jurisdictions recently agreed on a ‘global tax reform’ in the OECD Inclusive Framework to be effective in 2023,[xv] it can be reasonably assumed that the issues touched upon in this blogpost will likely become all the more relevant in the years ahead. Correspondingly, the need for profound research in this area is much-needed, making it both exciting and challenging times for tax scholars.
[i] Eg J Heinrich and H Moritz, ‘Interpretation of Tax Treaties’ (2000) 40(4) European Taxation 148.
[ii] Eg H J Ault, ‘The Role of the OECD Commentaries in the Interpretation of Tax Treaties’ (1994) 22(4) Intertax 146–147; K Vogel, ‘The Influence of the OECD Commentaries on Treaty Interpretation’ (2000) 54(12) Bulletin for International Fiscal Documentation 615; J F Avery Jones, ‘The Effect of Changes in the OECD Commentaries after a Treaty is Concluded’ (2002) 56(3) Bulletin for International Fiscal Documentation 102–103.
[iii] A noticeable deviation from this general practice can be identified in the Austrian tax treaty policy. Since the 1990s, Austria has been eager to implement respective interpretation clauses in its bilateral tax treaties. Eg D A Ward, ‘The Role of the Commentaries on the OECD Model in the Tax Treaty Interpretation Process’ (2006) 60(3) Bulletin for International Fiscal Documentation 100 (at fn 19).
[iv] For respective discussions, see eg Avery Jones (n 3) 103–104; P Wattel and O Marres, ‘The Legal Status of the OECD Commentary and Static or Ambulatory Interpretation of Tax Treaties’ (2003) 43(7) European Taxation 222–223.
[v] See for an overview over the case laws in selected countries Wattel and Marres (n 7) 229–232.
[vi] The beneficial ownership requirement is implemented in the allocation rules for dividends, interests and royalties (Art.10 para.4, Art.11 para.4 and Art.12 para.3 OECD MTC respectively). The foreseeable relevance is a requirement for legitimate information requests under Art.26 OECD MTC.
[vii] The following assessment is focused on the preliminary questions that are relevant for the topic of this blogpost.
[viii] For a critical discussion of the CJEU’s arguments, see S Geringer, ‘Implications of the OECD MTC Commentary’s dynamic interpretation in ECJ case law: the OECD as EU legislator?’, SSRN Working Paper, accessible via https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3979779 (last assessed 22 Dec. 2021), at sec. 4.1.
[ix] I.e. the OECD’s 1986 report on ‘Double Tax Conventions and the Use of Conduit Companies’; Geringer (n 11), at sec. 4.1.
[x] Joined Cases C-115/16, C-118/16, C-119/16 and C-299/16  N Luxembourg 1 and Others v Skatteministeriet EU:C:2019:134, paras. 92–93.
[xi] Recital 9 of the DAC.
[xii] Geringer (n 11), at sec. 4.1.
[xiii] In Cobelfret (C-138/09) the CJEU still took another stance; Geringer (n 11), at sec. 4.1.
[xiv] This concerns Bulgaria, Croatia, Cyprus, Malta and Romania; https://www.oecd.org/about (last assessed 22 Dec. 2021).
[xv] https://www.oecd.org/tax/beps/statement-on-a-two-pillar-solution-to-address-the-tax-challenges-arising-from-the-digitalisation-of-the-economy-october-2021.pdf (last assessed 22 Dec. 2021).