AX C ONVENTION AS AN INTERPRETATIVE AID : THE STATIC VS AMBULATORY APPROACH DEBATE CONSIDERED FROM A S OUTH A FRICAN PERSPECTIVE

Most tax treaties (including South Africa’s) are based on the OECD Model Tax Convention on Income and Capital and the related Commentary (the ‘OECD Model’). Notwithstanding the uncertainty surrounding its legal status, the courts in many countries use the Commentary in the interpretation of treaties. This article aims to contribute to the debate regarding the use of a static or ambulatory approach when using the OECD Model Commentary. If a double tax agreement (DTA) is based on the OECD Model and a certain provision follows the wording of the OECD Model, it could be contended that the contracting states intended such a provision to have the meaning it has in the OECD Model. However, the interpretation of revisions made to the OECD Model and Commentary subsequent to the conclusion of a DTA remains contentious, as scholars appear to be divided between an ambulatory and a static approach. A four step approach is recommended when considering the application of the Commentary


Introduction
The Organisation for Economic Cooperation and Development (OECD) aims to improve international tax co-operation between governments.This is partly in an attempt to counter international tax avoidance and evasion by identifying trends in international tax planning and helping governments to respond more quickly and effectively to emerging risks (OECD, 2015).In furtherance of these goals, the OECD launched an action plan on Base Erosion and Profit Shifting (BEPS) in 2013.The BEPS initiative was endorsed by the governments of the G20 countries and therefore extends its application to some OECD non-member countries.
BEPS refers to tax planning strategies that 'exploit gaps in the architecture of the international tax system to artificially shift profits to places where there is little or no economic activity or taxation' (OECD, 2015).In South Africa, the importance of combating BEPS is highlighted by the fact that the Davis Tax Committee (2014) has appointed a sub-committee specifically to address issues pertaining to BEPS.Moreover, due to the fact that the OECD's recommendations have become a globally accepted standard, together with South Africa's membership of the G20 community, an argument could be advanced that South Africa is bound, in a way, to follow BEPS (thereby escalating its importance).
With more than 3 000 bilateral tax treaties having been signed to date (Lang & Owens, 2014:6)of which South Africa has approximately 80 double tax agreements ('DTAs') in force (SARS, 2015) -the correct interpretation and application of tax treaties are viewed in a serious light by the South African National Treasury and its global counterparts.

Research objective
Notwithstanding that the OECD Model Tax Convention on Income and Capital 2014 and related Commentary (the 'OECD Model') can be of great assistance in the application and interpretation of tax treaties and in the settlement of disputes, it will be seen that their legal relevance remains a globally contentious point.As such, it is the primary objective of this article to contribute to the debate regarding the use of a static or ambulatory approach when using the OECD Model Commentary.In furtherance of this goal, a synthesis of scholarly opinions will be examined.

Research method
An interpretive research approach will be adopted for this study, as it seeks to understand and describe (Babbie & Mouton, 2009).As with most legal interpretive research, this study adopts a doctrinal research methodology, as it provides a systematic exposition of the rules governing a particular legal category (in this case, the legal rules pertaining to the OECD Model), explains areas of difficulty and is based purely on documentary data (McKerchar, 2008).This desktop study entails a literature review of and reference to both foreign and local statutory laws, tax treaties and policy documents, as well as authoritative studies on model tax conventions and double tax agreements.The documentary data to be used will be obtained from published articles, chapters in books, journal and legal databases and reputable websites.The research will reflect the law and policy developments up to and including 30 April 2016, except in certain circumstances where more recent policy developments or tax amendments appear particularly relevant.

Background
An earlier paper contains a concise background to the OECD and UN models, as well as an exposition as to why the OECD Model applies to a non-OECD member country such as South Africa (see Steenkamp, 2017).Accordingly, it is considered beyond the scope of this paper to again elaborate on these matters.
Despite not being a member of the OECD, most of South Africa's DTAs largely follow the OECD Model guidelines, as these are regarded as important and influential (Oguttu, 2007:242;Haupt, 2014:493).Suffice it to say that if the treaties of non-member countries are in conformity with the OECD Model and no specific position has been taken, the non-members also accept the provisions of the OECD Model and the Commentary as an interpretative aid (Wattel & Marres, 2003:224).
It has been the practice of the UN and the OECD that model tax conventions are all accompanied by Commentary notes, which are regularly updated approximately every two years (Ward, 2006:97).In recognition of the need to address new tax issues that arise in connection with the evolution of the global economy, the OECD released the contents of the 2014 update to the OECD Model on 16 July 2014.Previous updates were published in 1994, 1995, 1997, 2000, 2003, 2005, 2008 and 2010.While this recent update affects both the Articles of the OECD Model and the Commentary, most of the changes are in respect of the Commentary.This article is based on the 2014 OECD Model and Commentary.

Rules of interpretation from a South African perspective
The Constitution of the Republic of South Africa 1996 deals with international agreements in s 231, with international customary law in s 232 and with the application of international law in s 233.Olivier and Honiball (2011:303) state that, as a treaty is classified as an international agreement, it has to be applied in accordance with s 231 of the Constitution.Section 232 of the Constitution provides that international customary law is law in the Republic unless it is inconsistent with the Constitution or an Act of Parliament.
It is perhaps pertinent at this stage to point out that there are dissenting opinions about the status of the models (and their commentaries) as international law.Du Plessis (2012:52) provides an informative synthesis of scholarly opinion on the status of tax treaties in South African law and remarks that the South African courts have not explicitly pronounced on the status of the OECD Commentary and that they seem to refer to the OECD Commentary without providing reasons for doing so.On the premise that the models and commentaries are not international law, these will not form part of South African domestic law in terms of the Constitution (Du Plessis, 2012:44).
In a number of fairly recent decisions, the South African courts have been called upon to pronounce on the application of DTAs entered into by South Africa.One of the main points of contention regarding the interpretation of DTAs is whether tax treaties should be interpreted according to domestic or internationally accepted rules of interpretation.This dichotomy between domestic and international law, as well as the related matter of treaty override, falls well outside the scope of this article.For useful articles and leading textbooks from a South African perspective on these matters, see, for example, Clegg and Stretch (2015); Costa and Stack (2014); Du Plessis (2014); Du Plessis (2012); Olivier and Honiball (2011);Brincker (2010aBrincker ( , 2010b)); Hattingh (2010);and Oguttu (2007).
Suffice it to say that, notwithstanding the uncertainty surrounding the legal status of the OECD Model and Commentary, the courts in many countries use the OECD Model Commentary in the interpretation of treaties (Baker, 2002:para E.10; OECD Model, Introduction:para 29.1), although the exact basis on which they do so is unclear and not frequently explicitly stated by the courts (Baker, 2002:para E.12;Erasmus-Koen & Douma, 2007:349).The OECD itself intends for the Commentary to be used in the interpretation of DTAs (OECD Model, Introduction: para 29).
The weight of academic analysis in South Africa falls on the side of the Commentary not being legally binding.Nevertheless, the Commentary is a highly influential tool for the courts and can offer significant assistance for interpreting DTAs that follow the OECD Model.Olivier and Honiball (2011:321) aver that the OECD Model Commentary probably forms part of South Africa's customary international law on the basis of its acceptance in South African case law.They therefore consider the OECD Model Commentary to be relevant in interpreting treaty provisions in South Africa.
A significant point of ambiguity regarding the use of the OECD Model Commentary is the version that ought to be used in the interpretation process.This is the subject matter of the next paragraph.

Introduction
If a DTA was concluded in a given year and the OECD subsequently amended the Commentary, which version of the Commentary is to be used when a dispute arises?The OECD Committee on Fiscal Affairs recognised that the model had to be updated on an ongoing basis and, accordingly, adopted the concept of an ambulatory model (Olivier & Honiball, 2011:270).This means that specific articles in the model are adapted on a more regular basis without waiting for a complete revision.An ambulatory interpretation provides that the law at the time the treaty is applied has to prevail, and not the law at the time that the treaty was entered into.
Proponents of the static approach uphold the arguments of legal certainty and the pacta sunt servanda rule (meaning 'agreements must be kept'), in terms of which a DTA should be interpreted based on the intentions of the parties at the time it was entered into (Olivier & Honiball, 2011:301).The authors explain that subsequent developments are considered irrelevant, as the parties did not consider them at the time the treaty was entered into (and neither did Parliament).Supporters of the ambulatory approach argue that the failure to adapt tax laws to an everchanging environment may result in a situation which was never intended by the contracting states (Olivier & Honiball, 2011:301).Moreover, a treaty interpretation based on references to provisions or assumptions which are obsolete, no longer applicable or no longer permissible, may be 'extremely laborious' (Wattel & Marres, 2003:223).
Irrespective of the approach preferred, a two-fold interpretational problem exists: firstly, posttreaty changes in national law where the treaty refers to national law and, secondly, post-treaty changes in the Commentary to the OECD Model on which the treaty is based.This two-fold interpretational problem will be addressed below.

Changes in national law
As regards the first dilemma, another question appears: should the national law at the time the treaty was concluded or subsequently applied be consulted?Rocha (2012:358) suggests that the question regarding the adoption of a static or ambulatory approach to domestic law should be addressed in a tax treaty itself.In the absence of such a rule in the tax treaty, it should be for the interpreter to determine which legislation to consider (Rocha, 2012:358).
It could be argued that, when a treaty refers to national law for its interpretation, ambulatory interpretation must be the starting point, bearing in mind the context of the provisions and the 'good faith' provision of Art 31 of the Vienna Convention on the Law of Treaties 1969 ('VCLT') (Wattel & Marres, 2003:223).In light of the good faith provision, it could be argued that it is for the contracting states to choose between the law in force at the time of the conclusion of a DTA and the law in force at the time of the interpretation (Rocha, 2012:358).Moreover, the author is of the opinion that it is the duty of the contracting states not to adopt an interpretation that is contrary to the interests of both parties, as set out in that international agreement.

Changes in the OECD Model Commentary
In respect of the second dilemma, the question arises as to which version of the OECD Commentary (and, concomitantly, which version of the OECD Model) ought to be consulted.Paragraph 35 of the Introduction to the OECD Model seems to uphold a dynamic interpretation of treaties: Needless to say, amendments to the Articles of the Model Convention and changes to the Commentary 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.However, other changes or additions to the Commentary are normally applicable to the interpretation and application of conventions concluded before their adoption, because they reflect the consensus of the OECD member countries as to the proper interpretation of existing provisions and their application to specific situations.
Since 1992, the OECD has published both the Model and Commentary primarily as a loose-leaf collection.The amendments are published by delivering them as instalments to the subscribers of the loose-leaf.Vogel (2000:615) mentions a number of advantages (at least intended) by this change, amongst which are:  the Committee on Fiscal Affairs can react sooner to newly emerging problems that may be urgent;  additional reports that were published separately and statements of certain non-OECD countries are included in the loose-leaf publication; and  taxpayers, administrators and judges who have an edition of the loose-leaf binders (assuming that the binders have been diligently cared for and include the most recent instalments) can be sure that they will find in these two volumes all the information on the current Model and Commentary that they need.
Although aimed at making the Model more topical, more dynamic and facilitating the ease of working with it, the loose-leaf format also has a number of drawbacks (Vogel, 2000:615):  the reliability thereof has a tendency to decrease, very often as soon as after a few instalments;  even the most scrupulous person makes mistakes when inserting pages; and  even the OECD instalments are not absolutely reliable (e.g.severe printing errors and sentences which had been omitted).
It could be advocated that the OECD includes the Commentary and their new versions in a type of 'official' journal of the OECD, as this would improve legal certainty and aid in identifying which version treaty negotiators may have followed when negotiating a treaty (Martín Jiménéz, 2004:29).
As previously stated, the OECD's view is that existing treaties should be interpreted in the spirit of the revised Commentary, as far as possible (OECD Model, Introduction:para 33).However, if the revised articles or Commentary differ in substance from those used in previously concluded treaties, the OECD concedes that the revised Commentary is irrelevant.
The OECD's stance that an ambulatory approach is to be applied is not shared by all.In contrast to the above scenario, Wattel and Marres (2003:224) prefer a static interpretation as the starting point for changes in the Commentary.Their argument is that the Commentary current at the time of treaty conclusion can be considered to have received parliamentary approval and thus have become both part of the context of the treaty and parliamentary history.Lang and Brugger (2008:106) concur, explaining that the relevance of the OECD Model Commentary in the interpretation of tax treaties is based on the assumption that the contracting states, by following the wording of the OECD Model in drafting a certain provision, intended such a provision to have the meaning it has in the OECD Model, as outlined in the OECD Model Commentary.The relevance of a particular Commentary version thus depends on whether this assumption can be maintained (Lang & Brugger 2008:106).Sharkey (2001:658) holds an opposing view and prefers an ambulatory interpretation.His reasons are as follows: the taxes dealt with in tax treaties change rapidly; if an ambulatory approach is not followed, treaties will become less useful.He uses Art 2 (Taxes covered) of the OECD Model as a case in point: if the term 'existing taxes covered' is not given an ambulatory approach, it could be argued that, as tax changes over time and new developments in tax arise which are not covered by the tax treaty, this could defeat the purpose of the DTA (Sharkey, 2001:658).Elliffe (2013:36) also offers arguments in support of an ambulatory approach: first, given the process of treaty negotiation, where a typical DTA might take two years to negotiate, but some may take more than a decade, negotiators are aware of and utilise OECD positions that may not yet be public.Secondly, the OECD Commentary also reflects an international organisation's view that is officially approved by the governments of OECD member countries and not the view of any one particular tax administration that may be a party to the particular dispute (Elliffe, 2013:36).A court therefore does not fail to interpret a tax treaty fairly if it refers to the OECD Commentary.
The dynamic approach provides a solution to an important practical problem: if the articles of the OECD Model are modified, it would take years before the changes are generally included in DTAs worldwide, as existing treaties would have to be renegotiated (Martín Jiménéz, 2004:28).Vogel analyses the position held by a delegate to the Committee on Fiscal Affairs (which appears to be shared by other members) as follows (Vogel, 2000:612): The reason for this statement [Changing the OECD Commentary is preferable to changing the Model Convention] ... was his consideration that changes in the Commentary (or better, Commentaries) become effective immediately by their adoption, whereas a change in the Model Convention will not become effective before existing treaties are modified or new treaties concluded in consequence of the change, both of which require tedious procedures.
As regards the OECD Commentary which existed at the time the DTA was concluded, West (2009:29) observes that most scholars appear to be in agreement that the OECD Commentary updates fall within Art 31 of the VCLT.Moreover, the Commentary formed part of the legal context of the tax treaty and can be presumed to reflect the intended interpretation of the tax treaty articles that follow the OECD Model (Gusmeroli, 2010:204).Consequently, later Commentary amendments cannot serve to establish the parties' intentions upon conclusion of a DTA (Lang & Brugger, 2008:107).Ward (2006:102) agrees, suggesting that the new Commentary has no legitimate role to play in the interpretation of DTAs made between the OECD (or even non-OECD) countries and concluded before the reversal of the prior Commentary.
One approach is to consider Commentary sections which have remained unchanged throughout the negotiation and ratification process in the interpretation of a DTA.This line of reasoning would imply that Commentary amendments adopted after the ratification of a DTA may be taken into account only in exceptional cases (Lang & Brugger, 2008:106).The authors comment that it would be a difficult matter, absent any consensus of the contracting states, to regard a Commentary after ratification in the same way as a Commentary before, if only because the revised Commentary was not taken into account by the parties to the treaty before adopting the particular provision (Lang & Brugger, 2008:106).Vogel (2000:612) reasons that changing the OECD Commentary is preferable to changing the OECD Model itself, as changes to the former become effective immediately by their adoption, whereas changes to the latter will not become effective before existing treaties are modified or new treaties are concluded in consequence of the change (both of which require tedious procedures).Lang and Brugger (2008:102) take the following position regarding later Commentary amendments:  These shed no light on the intentions of the contracting states upon conclusion of the DTA. They are not part of the context (as defined in Art 31 (2) of the VCLT).
 They may play a role under s 31(3), which refers to 'any subsequent agreement between the parties regarding the interpretation of the treaty or the application of its provisions' and 'any subsequent practice in the application of the treaty which establishes the agreement of the parties regarding its interpretation'.
Article 31(3)(a) of the VCLT deals with the issue of a 'subsequent agreement'.Such an agreement must be binding on the parties under international law (Lang & Brugger, 2008:104).Since neither the OECD Model nor the Commentary is a legally binding instrument, an amendment to the Commentary does not constitute an agreement under international law and thus falls outside the scope of Art 31(3)(a) (Lang & Brugger, 2008:105).
Article 31(3)(b) of the VCLT deals with the concept of 'subsequent practice' which occurs when the tax administrations of both contracting states consistently apply an interpretation introduced through a Commentary amendment; such an amendment may become relevant through subsequent practice (Lang and Brugger, 2008:103).This entails that the parties' current understanding of the DTA (and not just their original intent) is held to be relevant; however, this practice may be applied only to clarify an otherwise ambiguous interpretation result (Lang & Brugger, 2008:104).
Although the OECD engages in discourse with the general public on certain topics and representatives of private companies participate in the OECD working groups, it could also be advisable for the OECD to add a countervailing weight to the opinions of tax administrations (Martín Jiménéz, 2004:29).This could be done by setting up a consultation group consisting of experts on treaties, who are ideally not connected with a tax administration, which could release opinions on the interpretation of treaties and revisions of the Commentary (Martín Jiménéz, 2004:29).
Notwithstanding that the Commentary may be taken into account in interpreting treaties, it remains doubtful whether a static or ambulatory approach should be followed (Olivier & Honiball, 2011:314).At the very least, the Commentary updates provide an insight into the interpretation of DTAs.West (2009:31) submits that some of the later Commentary updates should be considered to also have a persuasive effect, but that the use of the later Commentary should be decided on a case-by-case basis.This study concurs that the Commentary has persuasive effect, but it is furthermore submitted that caution should be exercised with regard to the application to DTAs which pre-date the specific set of Commentary updates.If one takes the view that later modifications to the Commentary affect prior treaties in that they should be interpreted according to the new Commentary, this could also present problems in terms of taxpayers' rights and constitutional principles.

CONCLUSION
As is the case with DTAs, the OECD Model Commentary is the result of compromise, reflecting observations and reservations made by states.They are generally phrased using fairly flexible language so as to accommodate a wide range of opinions (Linderfalk & Hilling, 2014:15).Consequently, the precise meaning of the Commentary will likely remain a subject of considerable debate.
If a DTA is based on the OECD Model (as most of South Africa's DTAs are) and a certain provision follows the wording of the OECD Model, it could be contended that the contracting states intended such a provision to have the meaning it has in the OECD Model.However, the interpretation of revisions made to the OECD Model and Commentary subsequent to the conclusion of a DTA remains contentious, as scholars appear to be divided between an ambulatory and a static approach.This article recommends Vogel's (2000:616) four-step approach when considering the application of the Commentary:  If a term was already used in the original 1963 Model and explained by its Commentary, it should be considered to have become in the course of time part of the 'international tax language'.As such, it should be assumed that the meaning attributed to that term is its ordinary meaning within the ambit of Art 31(1) of the VCLT.
 If a term was used for the first time in a revised Model, the meaning attributed to it may also have become its ordinary meaning in the interim.However, this is doubtful if the DTA was concluded during the first years after the introduction of the revised model.
 If the meaning ascribed to a term was not the ordinary meaning, it should be determined whether there is a special meaning within the ambit of Art 31(4) of the VCLT.However, this is justified only if enough time has elapsed between the amendment of the Commentary and the conclusion of the particular DTA.
 If the meaning is too recent, the Commentary may still serve as a supplementary means of interpretation, subject to the limitations of Art 32 of the VCLT.
It would certainly be helpful if the South African courts took notice of the uncertainty surrounding the basis on which courts may refer to the Commentary and provided guidance on these points.In any event, it is hoped that this article contributed to the debate by illuminating some aspects regarding the use of a static or ambulatory approach.
The author also refers to a classification by the OECD Working Party 1 as to the use of later Commentary:  filling a gap in existing Commentary by covering matters not discussed at all;  amplifying existing Commentary by adding new examples or arguments supplementing what is already there;  recording what states have been doing in practice; and  contradicting previous Commentary.