15 April 2021

OECD commentaries – how they affect interpretation of double tax treaties after adoption

Maarten van der Weijden

(This article was written in collaboration with our Best Friends colleagues Yves Rutschmann and Victor Camatta at Bredin Prat, France). OECD commentaries provide tax authorities, taxpayers and judges with key insights on how double tax treaties should be applied. But if recommendations change after a treaty has been adopted, their impact is less clear-cut and OECD member countries take different positions on this.

Two recent decisions by the French Administrative Supreme Court and the Amsterdam Court of Appeal highlight the issue. As evidenced by these decisions, the courts should keep in mind that a double tax treaty is first and foremost an agreement between two states, and that the courts’ role is to apply the treaty in a manner consistent with the initial intentions of both contracting parties. This alone should prevent the unlimited, automatic use of subsequent OECD commentaries.

OECD’s non-binding recommendation

Reasserting an OECD Council recommendation dated 23 October 1997, paragraph 35 of the OECD Commentaries to the preamble to the OECD Model Tax Convention expressly states:

Needless to say, amendments to the Articles of the Model Convention and changes to the Commentaries that are a direct result of these amendments are not relevant to the interpretation or application of previously concluded conventions where the provisions of those conventions are different in substance from the amended Articles (see for instance, paragraph 4 of the Commentary on Article 5). However, other changes or additions to the Commentaries are normally applicable to the interpretation and application of conventions concluded before their adoption of these changes, because they reflect the consensus of the OECD member countries as to the proper interpretation of existing provisions and their application to specific situations”.

Different positions

However, this non-binding recommendation is not being followed by all OECD member countries, as illustrated by these examples:

  • Germany: the Federal Tax Court does not take into account OECD commentaries published after the adoption of the relevant double tax treaty;
  • Spain: the Spanish Supreme Court has traditionally adopted an ambivalent approach on the subject of “dynamic” versus “static” interpretation of tax treaties, but in 2020, it ruled that OECD commentaries “cannot be applied retroactively”;
  • UK: according to the UK Supreme Court in Fowler v HMRC (20 May 2020), OECD commentaries which post-date a particular double tax treaty should be given “such persuasive force as aids to interpretation as the cogency of their reasoning deserves”;
  • Italy: the Italian Supreme Court refuses to give decisive importance to the commentaries, but is willing to consider them as a comforting element of a prima facie interpretation of the treaty.

France and the Netherlands contribute to this list of varied positions, as illustrated by two recent, partly connected decisions.

French Administrative Supreme Court

In a decision dated 11 December 2020, this court ruled on the existence of a permanent establishment based on the “dependent agent” test of Article 2 of the 1968 French-Irish double tax treaty. At the heart of the case was the assessment of whether a French entity had, and had habitually exercised in France, the authority to conclude contracts in the name of an Irish group company and whether it was therefore acting in France as a dependent agent of the Irish entity.

The Paris Administrative Court of Appeal had previously ruled that the French entity did not have the authority to conclude such contracts. It held that even though the French entity’s employees were involved in the preparation, negotiation and implementation of the contracts between the Irish entity and its clients, these contracts (including any deviation from the Irish entity’s general terms of sale) could not legally bind the Irish entity without its prior approval and signature, even if this approval was “purely formal”.

The French Administrative Supreme Court overturned this ruling by holding that a French entity should be deemed to exercise authority to conclude contracts in the name of an Irish entity if the French entity habitually decides transactions which the Irish company merely ratifies and which, once ratified, are binding on the Irish company even though the French company does not formally conclude contracts in the Irish company’s name. To support this “substance over form” interpretation, the French Administrative Supreme Court expressly referred to OECD commentaries C(5) no. 32.1 and 33. These commentaries were published in 2003 and 2005, after the adoption of the 1968 French-Irish treaty, and supported this broad interpretation.

The use of subsequent OECD commentaries: the confirmation of a trend rather than a “revolution”

In its opinion provided to the French Administrative Supreme Court, the rapporteur public (to a certain extent equivalent to an advocate general) compared how the question of use of subsequent OECD commentaries is addressed by other OECD countries. This approach is not new and shows the will of the French Administrative Supreme Court, especially for new questions of law, to consider solutions applied abroad before taking its own decision. This is all the more understandable in international tax matters like the case before the court.

By explicitly referring to OECD commentaries published after the adoption of the French-Irish treaty, this decision raises the question of whether it is consistent with the court’s SA Andritz decision on 30 December 2003 (n° 233894). In that decision, the court expressly set out the principle that OECD commentaries published after the adoption of the relevant provisions of a double tax treaty could not be used when interpreting and applying those provisions.

According to leading scholars and judges, this principle was not, however, absolute and did not prevent tax courts from giving “persuasive” or “comforting” value to subsequent OECD commentaries, especially when it came to provisions defining permanent establishments. In other words, although subsequent OECD commentaries cannot be used when they extend the scope of the relevant treaty provision as initially agreed by the contracting states, they may be used when they merely clarify the original meaning of the treaty provision without expanding its scope.

Given the wording used by the 2020 decision (“as it can furthermore be seen from paragraphs 32.1 and 33 of the [OECD commentaries]”), the French Administrative Supreme Court followed this approach, as subsequent OECD commentaries seem to have been used only to comfort – as persuasive elements – its own interpretation of the concept of “authority to conclude” contracts within the meaning of the French-Irish treaty. In the case at hand, it means that the French Administrative Supreme Court implicitly considered that the concept of authority to conclude contracts, as initially agreed by the contracting parties and subsequently confirmed by the OECD commentaries, had to be construed as covering instances where the foreign entity’s signature and approval of the transaction as decided and negotiated by the agent are only formal and given on a routine basis.

This decision should therefore not be construed as introducing a new and absolute exception to the principle set out in the 2003 SA Andritz decision, but merely as explicitly confirming for the first time the potential persuasive nature of subsequent OECD commentaries. This solution is understandable as it would be odd to generally exclude subsequent OECD commentaries irrespective of their content.

A new era of uncertainty or an opportunity for taxpayers?

Determining when later OECD commentaries merely clarify a treaty provision or actually modify it on a case-by-case basis will not be easy in practice. This “case-by-case” approach inevitably creates some uncertainty for the taxpayer, but it may also provide some opportunities when subsequent OECD commentaries give an interpretation more favourable than the one opposed by the tax authorities.

Amsterdam Court of Appeal

This decision (dated 22 December 2020) addressed whether OECD commentaries published after the adoption of the 1973 double tax treaty between France and the Netherlands could be taken into account when applying the treaty.

The facts of the case

The case involved the acquisition of a Dutch company by a French one via a tender offer. Following completion of the offer, the French company allocated the shares it had acquired in the Dutch company to a Dutch permanent establishment. The French company incurred costs in connection with the issue of shares and other equity instruments and claimed a deduction for these costs in the Dutch corporate income tax return of its Dutch permanent establishment. Under Dutch domestic law, there is clear case law that indicates that the costs involved in raising equity for the acquisition of the Dutch company constitute costs incurred as a result of the French company’s legal form (orgaankosten). These cannot be allocated to the income from the acquired shares in the Dutch company. As such, they cannot be claimed as a deduction within the Dutch permanent establishment.

Court of appeal’s reasoning and decision

The court of appeal considered, among other things, whether the French-Dutch treaty overrides Dutch tax law. The court first established that the text of the treaty did not unequivocally provide that the costs should be attributed to the Dutch permanent establishment and that the Dutch official explanation to the treaty did not discuss article 7 of the treaty or the allocation issues associated with that provision. The court then held that both treaty partners had referred to the OECD commentary to article 7 of the 1963 OECD Model Tax Convention for the interpretation of article 7 of the French-Dutch treaty. Because the text of article 7 of the French-Dutch Treaty is almost identical to the text of article 7 of the 1963 OECD Model Tax Convention, the court held that the commentary to that model could serve as an additional means of interpretation within the meaning of article 32 of the 1969 Vienna Convention on the Law of Treaties. The court also referred to a 1992 ruling of the Dutch Supreme Court that OECD Commentaries are “of great relevance” for the interpretation of treaties which are based on the OECD model tax convention. The court analysed the OECD commentaries on article 7, paragraph 2 of the 1963 OECD Model Tax Convention, in particular sub-paragraphs 12-14, and concluded that the commentaries did not support attributing the costs incurred by the French company to the Dutch permanent establishment.

The court of appeal then went on to consider whether the commentary on the 2010 OECD Model Tax Convention could be relevant in interpreting article 7 of the French-Dutch treaty. The court held that these commentaries would not necessarily be “of great relevance” because the text of the treaty provision had changed. The court referred, for what it called the “limited dynamic interpretation”, to paragraph 35 of the OECD commentary on the preamble of the 2010 Model Tax Convention, which in substance read the same as paragraph 35 cited above. The court also referred to a 1998 Supreme Court decision and held that the 2010 OECD Model Tax Convention changed the text of article 7 of the model by introducing the “authorised OECD approach”. Under this approach, the methodology followed to attribute profits to a permanent establishment differed from the methodology that was prevalent under the 1963 OECD Model Tax Convention. Therefore, the OECD commentary on the 2010 convention could not be used to interpret article 7 of the French-Dutch treaty; doing so would be in conflict with the “limited dynamic approach”, according to the court.

The court of appeal essentially decided that, with the changed text of article 7 of the OECD Model Tax Convention and the commentary, the OECD had crossed the line of being “different in substance” as referred to in paragraph 35 of the OECD commentary to the preamble. Although the authorised OECD approach admittedly changed the methodology for the attribution of profits to a permanent establishment, whether this change was supposed to have an effect on the allocation of costs such as the ones at stake in this case, was not certain.

What will the Dutch Supreme Court rule?

The French taxpayer involved has lodged an appeal against the court of appeal’s decision with the Dutch Supreme Court. How the court of appeal set up its reasoning, including specific references to prior Dutch Supreme Court decisions, sets the stage for a principled decision by the Dutch Supreme Court on how to deal, in this time and age where we have seen an expanded OECD role, with changes to provisions of the OECD Model Tax Convention and OECD commentaries published after a treaty between the Netherlands and a third state has been concluded.

More to come

As the OECD’s role is setting tax policy and taking the lead in the implementation of that policy in tax treaties, the question of how much weight should be given to the OECD commentaries is likely to gain importance. Subsequent OECD commentaries may definitively be seen as a useful interpretation tool, as they allow a dynamic interpretation of the treaty in question, which is more than relevant in an ever-changing world. However, and as evidenced by these two decisions, the courts should keep in mind that a double tax treaty is first and foremost an agreement between two states, and that the courts’ role is to apply the treaty in a manner consistent with the initial intentions of both contracting parties. This alone should prevent the unlimited, automatic use of subsequent OECD commentaries.