INTRODUCTION
This paper critically analyzes the emergence of a multilateral tax regime, particularly due to the recent efforts produced by the OECD. It appraises the meaning and potential impact of the initiatives adopted within this international organization to promote a multilateral approach to taxation within international law. The shift would be prompted by the need to overcome the limits of double taxation treaties, the urge to fight against base erosion and profit shifting (BEPS), and the necessity to adjust to globalization and the digital economy. It further contends that there is hardly anything exceptional about taxation that might warrant special treatment to this special field, therefore such emerging multilateralism should be carried out within the existing framework of international law.
International lawyers have all come to accept that “tax is different”, thus leaving it to be dealt with by international organizations dedicated to tax issues, mainly the Organization for Economic Cooperation and Development (OECD), instead of the main venues of international law, such as the United Nations (UN). The isolation of tax can also be demonstrated by its development vis-a-vis globalization. While many areas of law have
reacted positively to a perceived more global and interconnected society, tax remained
secluded, creating barriers against its ‘internationalization’.
For instance, not a bit more than two decades ago, in the 1998 Tillinghast Lecture delivered at the New York University Law School, it was stated that the existence of overarching principles of international taxation […], qualifies as news”(ROSENBLOOM, 1999, p. 137). In that same year, it was also argued that, in a strict sense there is no international taxation. All taxes great and small are creatures of purely national tax laws (ISENBERGH, 2006).
However, a lot has changed in the last two decades and international tax has never been as central in the world’s political and institutional agendas as it is today. Leaving taxes as a creature of domestic laws, without supranational coordination and consistency, has paved the way for the making of opportunities for tax arbitrage, leading to tax avoidance or evasion through the use of structures or strategies that exploit loopholes in tax rules, thus artificially shifting profits to low or no-tax jurisdictions where little or no real economic activity take place.
In order to fight against the use and implementation of such strategies, a package of measures has been developed by the OECD / G201, the so-called BEPS, an initiative that represents the first substantial – and overdue – renovation of international tax standards in almost a century. It is an important subject for the entire international legal community, not only for developed and industrialized nations but also for developing ones. At the same time, it represents the recognition by governments that they have been losing significant tax revenues due to aggressive international tax planning and that international / supranational coordination is necessary. As such, the OECD acknowledged that it was important to consider innovative, coordinated and consistent ways to implement, in a multilateral context, the measures that would result from BEPS and as such proposed the development of a multilateral instrument; the ‘Multilateral Convention to Implement Tax Treaty Related Measures to Prevent Base Erosion and Profit Shifting’ (MLI).
Being the first multilateral tax treaty covering significant substantive international tax obligations, the MLI offers concrete solutions to fend-off some of the gaps in existing international tax rules by transposing results from BEPS into the existing network of bilateral treaties that have been concluded to eliminate double taxation (DTT). The MLI also aims to alter and modify the application of such DTTs while at the same time implementing agreed minimum standards to counter treaty abuse and to improve dispute resolution mechanisms. Being it a novelty, it would be important to explore and analyze some of its policy choices to understand the new grounds of international tax.
This is a forum attended by finance ministers and central bank officials from the world’s highly developed economies consisting of 19 countries and the European Union.
While the MLI can clearly achieve some of its intended objectives, it will also pose and create new challenges to the international legal community. Moreover, it shall expose the field of international tax to the same challenges that have been terrifying international lawyers for ages: the lack of enforcement mechanisms, the legitimacy of international institutions and organizations, its ‘double-standard’, which allows rich and powerful states to act with impunity while poor ones are held accountable, debates about “soft” and “hard” law, “fragmentation”, among others. It will not be within the scope of this paper to delve into each and every one of these questions – an out-of-proportion task – on the contrary, this paper will simply try to raise some level of awareness regarding the so-called tax isolationism / exceptionalism and also discuss the emerging multilateralism that international tax is experiencing. The main argument remains quite simple: there is nothing inherently sovereign2 about taxes and or taxation.
The analysis herein will be done using concepts and experiences of other areas of international law, thus ending – or helping to end – a common understanding that tax is special, different, or exceptional. Since there is nothing special about taxation, it can – and should – be studied using knowledge gained by such other areas: trade, investment, environment, and human rights. The goal is to use public international law to guide the work and to assert that the so-called tax exceptionalism is a ‘creature of the past’ with the future presenting a handful of opportunities for cross-fertilization. In order to further accomplish that, it is mandatory to abandon certain old ‘dogmas’, thus embracing the understanding that tax is a legitimate concern of all people. Furthermore, in a globalized world, tax should be a central part of the international legal community’s challenges, debates, and efforts.
The paper is structured as follows: chapter 2 will start with a historical analysis of the international tax system in order to deconstrue, once and for all, the wrong assumption that tax is special or different and therefore positing that international tax is part of international law. Then, chapter 3 will look carefully into the MLI to try to define, within the boundaries of international law, some of its possible shortcomings. This will include its (limited) scope of application, general rules of interpretation and the policy choices it adopted regarding the improvement of dispute settlement mechanisms and if such choices are indeed adequate for the expected growing conflicts and disputes international tax law may be about to face. Chapter 4 concludes.
INTERNATIONAL TAX AS INTERNATIONAL LAW
Is international tax part of international law? That question, which seems quite odd and out of place today, was put by Avi-Yonah (2007) not too long ago. According to him, the
The treatment of citizens by governments within the boundaries of a given territory was once considered to be a matter of domestic jurisdiction, but it came to be covered by international human rights laws and principles. That change in perspective and attitude can be transposed and applied to international tax as well.
answer to such question was probably obvious to an international lawyer, who would undoubtfully answer that international tax is indeed part of international law. But to an international tax lawyer, that same question probably seemed less obvious, because most international tax lawyers did not think of themselves primarily as international lawyers, but rather as tax lawyers who happen to deal with cross-border transactions.
The reasons behind the perception that international tax was not part of international law could be explained by many reasons. Clearly there is a lingering assumption that taxation is absolutely sovereign and therefore cannot be dealt with at the international level. Another possible explanation, which seems informed by circumstances, is the fact that the international tax system has been characterized and formed by a multitude of bilateral tax treaties, such treaties not posing the same sort of questions and challenges that mark the interpretation and application of multilateral treaties.
The historic objective of the international tax system has been mainly geared towards addressing and tackling double taxation, thus preventing Multi-National Enterprises (MNE) from being taxed in every jurisdiction they had a presence at or an activity in. Such objective has been achieved by bilateral DTTs3, usually limited in scope and adopted without global coordination and harmonization. DTTs are usually the result of treaty policies and bargaining powers of the parties to them.
However, it seems undeniable that a DTT is a treaty for the purposes of the Vienna Convention on the Law of Treaties (VCLT). However, the VCLT is rarely relied on or invoked by during tax treaty disputes4. This lack of reliance on the VCLT further demonstrates the isolationism that has marked international tax and can be explained by the fact that most international tax courses do not teach the basics of international law as part of their introductory and mandatory basic curricula5. Be as it may, this remains a puzzling and uncomfortable fact which only exacerbates how tax has been perceived as exceptional and therefore demanding specialized knowledge, thus running parallel and in isolation to other areas of international law.
Things have started to change, and the limits of a system based on a network of bilateral treaties became more and more evident. Whilst it is undoubtful that DTTs have been a successful tool to fight double taxation, the progression of uncoordinated practices and harmful competition by States (both to attract investments and to raise funds) started to
Such treaties usually follow either the OECD Model Tax Convention (OECD, 2017) or the United Nations Model Convention (United Nations Model Double Taxation Convention between Developed and Developing Countries, 2013) For additional information on DTTs, see chapter 3 below.
According to Avi-Yonah (2007), in 2004 a search of the tax cases database in LEXIS would reveal among hundreds of treaty interpretation cases only one in which a US court had discussed the potential application of the VCLT (Kappus v. Commissioner, 337 F.3d 1053 (DC Cir. 2003), 2003).
Another possible explanation is the fact that although the United States has signed the VCLT, it has never ratified it. However, this explanation does not seem convincing as the experience in other jurisdiction seem to confirm that the VCLT in not part of the basic teachings in international law courses and it is also rarely raised as an argument in cases dealing with international tax disputes.
create other international tax issues – most important of them being double non-taxation, base erosion and profit shifting. Globalization, the digital economy and stronger capital mobility have just added more fuel to the ‘burning problem’, consequently demanding for a swift, coordinated, consistent and broader set of supranational rules to be developed, implemented, and applied thereto. The pressing issue became to stop or prevent the use of structures and strategies that exploit vulnerabilities and weaknesses of the – uncoordinated – system to artificially shift profits to low or no-tax jurisdictions where there is little or no real economic activity. The solution found is BEPS, which has been theoretically conceived to streamline taxation and to fight against tax arbitrage, avoidance, and evasion.
BEPS has set certain general recommendations, established some agreed minimum standards on treaty abuse and demanded for more effective dispute resolution mechanisms for tax treaty disputes. Accordingly, a multilateral context has emerged in international tax – probably not as a policy choice, but as a “necessary evil” – due to the need to find pragmatic solutions to transpose BEPS results into the existing network of bilateral DTTs. The MLI was the response to make the required amendments and changes to thousands of bilateral tax treaties possible, faster, and more streamlined. But being a multilateral treaty, it will undoubtfully create and pose other difficulties: reservations, objections to reservations, and amendments6; its legal relationship with existing DTTs, how it will regulate bilateral treaties to be concluded after its entry into force, and the use of the COP regarding its interpretation and implementation, just to name a few. At the same time, these very same examples help to demonstrate the expected greater interaction between international tax and international law and as such they should suffice to predict that this relationship will be as strong and intertwined as never before.
Therefore, the MLI provides the grounds for abandoning once and for all the old mainstream views that international tax is simply domestic tax which deals with cross- border transactions. Not only an international tax system does exist, but it should also be treated as an integral part of international law. As a result, both the UN and the OECD Model Conventions should be interpreted pursuant to the VCLT rules.7 Moreover, tax rules
– be it domestic or international – should be interpreted in light of principles that have informed the development of other areas of international law, including principles regarding the protection of citizens under international human rights. Finally, the UN should assume a more important, pivotal, and leading role in setting international tax
If unanimous such amendments are difficult to achieve and if by majority vote, create the subsequent issue of objection and dissent.
It is the author’s view that the VCLT could be enforceable even against countries that have not signed it as some of the rules it contains can be construed as a codification of customary international law.
principles and standards, thus conferring to them more legitimacy and effective representation8.
FROM BILATERAL TO MULTILATERAL TAX TREATIES
The international tax system can be described as being mainly based on the OECD Model Convention9 – which is currently the backbone of more than 3,000 bilateral tax treaties – the overwhelming majority of them comprised of income tax treaties.10 The other important supranational model convention is the UN Model Convention11, originally created for tax treaties between developed and developing countries. However, after careful examination, it can be stated that in fact, the UN Model Convention follows the OECD pattern with a lot of their provisions being identical, or nearly so, in such a way that it makes sense not to view the UN Model Convention as entirely separate and independent, but rather as making important, but limited, modifications to the OECD Model12. Moreover, the UN Model Convention had an exceedingly small effect on the conclusion of DTTs13.
The success of the Model Conventions, in its bilateral form, is a positive lesson the world needs to embrace. Its basic premise is the proper allocation of taxing rights between countries, known as the ‘benefits principle’14. Under the benefits principle, operating income (usually referred to as ‘active income’) should be taxed primarily at source while investment income (usually in the form of dividends, interest or royalties, usually referred to as ‘passive income’) should be taxed primarily at residence. However, since the 1980’s, tax competition has led many source jurisdictions to create Preferential Tax Regimes (PTR) in order to attract foreign direct investments, while at the same time residence jurisdictions have become reluctant to tax, thus not putting MNEs at a competitive disadvantage (AVI-
The UN system is not immune of having problems and deficiencies of its own, as the right of veto given to five states in the Security Council clearly demonstrates. However, it seems a more legitimate international organization than one made of a group of deve loped nations such as the OECD.
The first OECD Model Convention was published in 1958. The OECD Model Convention has been revised several times, most recently in 2017 (OECD, 2017).
There are other forms of international tax treaties, such as treaties dealing with inheritance or donation taxes for instance, but such treaties are quantitively extremely limited.
The origins of the UN Model Convention can be traced back to 1931, when the first effort to conclude a multilateral treaty took place. Such effort is represented by the reach by the subcommittee of the fiscal committee of the League of Nations of a draft that served later as the almost uncontested cornerstone for the analysis and research of international tax policies and rules.
The OECD Model Convention favors capital-exporting countries over capital-importing ones. Therefore, the OECD Model Convention may not be appropriate for treaties entered into by net capital-importing countries. As a result, developing countries devised their own model under the auspices of the United Nations. The main difference between the two model conventions is that the UN Model imposes fewer restrictions on the taxing rights of the source country; source countries, therefore, have greater taxing rights under it compared to the OECD Model Convention.
The UN Model Convention was last updated in 2001. Such model has had an influence on tax treaty negotiations between developing countries, but in tax matters the UN as a global institution has extraordinarily little influence. There may be many reasons for this, one being the prominence gained by the OECD itself, another could be the fact that the UN has not focused on, nor devoted resources to, tax issues. For instance, the UN has no secretariat or a fixed body of experts capable of making a difference in the field of tax coordination and cooperation, as it has in its other fields of interest. Finally, the very same tax exceptionalism addressed herein, could also be an explanation, as historically tax and international law have not met ‘under the same roof’.
On the benefits principle and its origins, see (AVI-YONAH, 2013).
YONAH; XU, 2017). The result has been that most MNEs are neither taxed at source nor at residence – the so-called double non-taxation.
In order to preserve income tax and to avoid double non-taxation in a globalized and more interconnected world, supranational solutions became needed, since “global challenges need global solutions”. Accordingly, one of these solutions could be the growing requirement to have the bilateral network of treaties replaced by a more coordinated endeavor in the form of a multilateral system – probably following the successful experiences in other areas of international law, such as the one represented by the World Trade Organization (WTO).
The pressing issues currently faced by the WTO and its Dispute Settlement Body (DSB)15 shows that multilateralism has its own perils and issues, but it is beyond the scope of this paper to conclude as to the appropriateness or desirability of the emerging multilateralism in the context of international taxes. What is clear is that social and economic conditions will keep challenging the way in which the system operates and as a result it is quite likely that taxes will become more complex, competitive and global (SAWYER, 2004) and, therefore, more coordination will be undoubtedly needed. One way such coordination can be achieved is by means of a stronger multilateralism. It is important to highlight that, although limited, multilateral tax treaties have already been used with successful implementation in a few cases of a well-defined region or in relation to administrative matters. Therefore, it is fair to say that today there are a few multilateral tax treaties currently in effect16.
On that topic, McIntyre argues that multilateral tax treaties’ political feasibility is dependent on the size of countries concluding them and their cultural similarities (i.e. the Nordic Convention, CARICOM) as well as its scope – i.e. the EU Arbitration Convention which only covers Transfer Pricing (TP) issues (MCINTYRE, 2002). Accordingly, a regional multilateral tax treaty seems to be more feasible than one that tries to encompass a large geographic base. In the same sense, a multilateral tax treaty will be more likely to be
At its November 2017 meeting the WTO DSB failed regarding an attempt to launch a selection process to fill the vacancies in the WTO Appellate Body. The United States withheld its agreement to launch the process. Appellate Body Members are appointed by the DSB, which comprises the representatives of all 164 WTO Members, DSB decisions are taken by consensus. The formal objection by any Member State can block appointment or re-appointment of Appellate Body Members. This could make the dispute settlement system of the WTO to break down According to the Chinese Ambassador to the WTO, Zhang Xiangchen the system is facing the most difficult time since its creation and “without such system, the WTO’s trade rules will no longer be effectively enforced, and the trust and credibility of the multilateral trading system will be deeply undermined.” (XINHUA, 2018).
The CARICOM Agreement (Agreement among the Governments of the Member States of the Caribbean Community for the Avoidance of Double Taxation and the Prevention of Fiscal Evasion with respect to Taxes on Income, Profits or Gains and Capital Gains and for the Encouragement of Regional Trade and Investment among Antigua and Barbuda, Jamaica, the Bahamas, Montserrat, Barbados, St. Kitts and Nevis, Belize, Saint Lucia, Dominica, St. Vincent and the Grenadines, Grenada, Suriname, Guyana, and Trinidad and Tobago ), the Andean Treaty (Regime for the Avoidance of Double Taxation and the Prevention of Fiscal Evasion among Bolivia, Colombia, Ecuador and Peru) the Nordic Convention (Convention between the Nordic Countries for the avoidance of double taxation with respect to taxes on income and on capital among Denmark, the Faeroe Islands, Finland, Iceland, Norway and Sweden), the Convention on the Elimination of Double Taxation in Connection with the Adjustment of Profits of Associated Enterprises (Convention 90/436/EEC of July 23, 1990), among others.
concluded if its scope is limited and if it only sets general and abstract overall standards, leaving sensitive issues to be dealt with through bilateral negotiations.
Regardless of the likelihood of its successful implementation, which can only be analyzed in years to come, the OECD has chosen a multilateral instrument as the solution for some of the challenges currently faced by international tax in general and BEPS in particular. However, before choosing the multilateral route, the OECD considered other options, such as an independent treaty that would comprehensively replace individual bilateral tax treaties, and a ‘one-time’ instrument which would operate as a collection of amending protocols, which would modify the wording of its underlying tax treaties (a ‘temporary’ instrument which would effectively lose its significance after implementation). While the first option was considered far too ambitious, the second was not chosen due to its perceived complexity and inefficiency (BOSMAN, 2017).
A third possible solution considered was the renegotiation of all existing tax treaties – a herculean task considering the 3,000 tax treaties currently in effect – through the publication of an updated version of the OECD Model Convention, that would be used and adopted as a basis for the required bilateral renegotiations. It happens that the OECD Model Convention is not binding, not even to OECD’s member-states17, thus there would be no guarantee that deviations from such model would not take place. Therefore, after considering all available options, based on the need to address BEPS in a comprehensive manner, the OECD recognized the need to contemplate innovative ways to implement the measures that would result from the project and, consequently, it determined the development of a multilateral instrument.
The Multilateral Tax Instrument
To design such instrument, more than one hundred jurisdictions participated in the negotiations, which culminated in the adoption of a final text, together with an explanatory statement, on 24 November 2016. The MLI entered into force on July 1st, 2018 in relation to its first five ratifying parties.18.
Since the MLI was not foreseen by existing DTTs, there is no direct relationship between them. Therefore, the MLI exists alongside and in addition to DTTs, in order to modify their application according to BEPS recommendations.19 As such, it works as an extra layer of
The OECD recommends that Member States follow the OECD Model Convention. That recommendation was as follows: “THE COUNCIL (…) RECOMMENDS the Governments of member countries (…) when concluding new bilateral conventions or revising existing bilateral conventions, to conform to the Model Tax Convention.” (OECD, [s.d.]).
18 Austria (9/22/2017), Isle of Man (10/25/2017), Jersey (12/15/2017), Poland (1/23/2018), and Slovenia (3/22/2018). Available at: < https://www.oecd.org/tax/treaties/beps-mli-signatories-and-parties.pdf>. Viewed on 20 jun. 2020.
While the MLI will constitute the first time a multilateral treaty will be used to modify bilateral tax treaties, this type of structure has already been used: the 2003 Agreement on Extradition between the EU and the United States. This 2003 agreement modified the provisions of existing bilateral extradition treaties between the different EU Member States and the United States.
treaty law, which is placed on top of existing DTTs. This entails that the MLI does not function as an amending protocol, which directly amends the wording of existing treaties, but rather it simply modifies them. Moreover, it was negotiated and drafted as to give as much flexibility as possible to the parties.
Except to certain provisions that reflect agreed minimum standards, the MLI is indeed quite flexible. It allows parties to choose between two or more provisions or opting in or out regarding optional ones. That is, parties can simply opt not to adopt a certain rule or to adopt the option that best fits its needs. Finally, the parties also have the choice to include existing DTTs within the scope of the MLI or not.
Reservations to certain provisions to the instrument can be made by notifying the depository (i.e. the OECD) of its intention. Accordingly, Article 28(1) allows a vast number of reservations, which can be operationalized in various and different ways and may take different forms. But it needs to be stated that all permitted reservations are expressly and duly contained in the treaty, hence they are limited in number.
Regarding the settlement of disputes, including conflicts of interpretation, the MLI opted for a Mutual Agreement Procedure (MAP) to be pursued in accordance with the rules contained in the DTTs. In case existing DTTs do not contain MAP provisions, as an agreed minimum standard, states must include such provisions thereto. However, it does not require the inclusion of mandatory and binding arbitration provisions and, if a state chooses to do so, strict confidentiality and secrecy is accorded.
Finally, as to questions and disputes which may arise out of interpretation and implementation issues, the parties to the MLI may use a Conference of the Parties (COP), which shall be convened in order to take any decisions or exercising any appropriate and relevant functions that may be needed or required.
An overall analysis of the MLI reveals that despite the fancy words and grand objectives, it fell short of putting forward strong proposals capable of establishing a new and solid foundation for a multilateral tax system capable of dealing and solving the difficult issues that might be ahead. Such shortcomings, although non-exhaustive, could be described as follows: (i) scope of its application, (ii) the making of reservations and notifications (iii) general rules of interpretation, (iv) improvement of the dispute settlement mechanism, more particularly the issue of mandatory arbitration, adjudication, and (v) the use of a COP. These issues will be individually analyzed in the next sub-chapters.
Scope of Application
While the MLI may be perceived as a pragmatic solution to the implementation of BEPS, being a tool to alter and modify existing bilateral tax treaties, its scope of application may be limited by the great flexibility attributed to it by its negotiators. Accordingly, the MLI
will only modify existing tax treaties which have been designated as “Covered Tax Agreements”. This means that a party can make an existing DTT not to be within the scope of the MLI by simply not listing such DTT as a Covered Tax Agreement.
Furthermore, other reasons could make the material scope of application of the MLI quite limited. First, a considerable number of states which took part in the negotiations have not yet signed it, including two of G20 countries: United States20 and Brazil21. Therefore, all tax treaties of these important countries will not be within the MLI scope of application.22 Second, nations who have signed and / or ratified the MLI may choose to leave some of their existing DTTs outside the realm of the MLI by simply not listing them as ‘Covered Tax Agreements’. Third, and finally, regarding provisions that are not considered to reflect an agreed minimum standard, such provisions are optional by means of certain allowed reservations. All these facts are extremely relevant to analyze the actual scope of the MLI (KLEIST, 2018).
Based on the preliminary notifications made during the signature of the MLI, the approaches taking by the signatories are quite diverse. Such diversity was described by Bosman as some states notifying nearly all their tax treaties (i.e. the Netherlands), while others taking a more conservative approach such as Switzerland (BOSMAN, 2017).
The issue of reservations, options and notifications is remarkably sensitive for the purposes of determining the actual scope and ultimate coverage of the MLI, as reservations allows for an almost infinite number of combinations. As Kleist (2018) properly argues the collective impact of such reservations is larger than may be suggested and careful attention should be devoted to this.
Based on the above propositions, if parties make too many reservations and opt out of several non-agreed minimum standards, the actual scope and material impact of the MLI will be extremely limited. Therefore, the complexity of the MLI is a sub-product of its flexibility that allows for an almost infinite number of combinations. The result of such many combinations is that the MLI will modify only a small percentage of existing DTTs and the scope of such modifications might become limited to a few core provisions that reflect agreed minimum standards.
MLI General Rules of Interpretation
Regarding interpretation, Article 2(2) of the MLI, provides for a general rule. According to such article, terms contained but not defined therein, unless the context requires
Based on the “US Treasury Official Comments on US Decision not to Sign MLI”, the United States may never sign the MLI (“Treasury Official on Why U.S. Did Not Sign BEPS Multilateral Instrument”, [s.d.]).
For a position on the Brazilian refusal to sign, see SCHOUERI (2008).
Nonetheless, this shall not mean that such countries will not implement BEPS’ agreed minimum standards by other means, for example through bilateral negotiations.
otherwise, shall have the meaning that it has under the relevant Covered Tax Agreement at the time the MLI is being applied. With respect to a term not defined in the MLI or in the relevant Covered Tax Agreement, the general rule of interpretation contained in Article 2(2) does not provide for a solution, being therefore silent. Notwithstanding such silence, according to the general rule of interpretation contained in Article 3(2) of the OECD Model Convention, which could apply in subsidiarity, “unless the context otherwise requires, undefined terms will have the meaning they have at the time the Covered Tax Agreement is being applied under the domestic law of the Contracting Jurisdiction applying the Covered Tax Agreement” (OECD, 2017).
According to Bosman, the provisions of Article 3(2) of the OECD Model Convention may even be “complemented with the general rules of treaty interpretation included in Articles 31 through 33 of the VCLT” (BOSMAN, 2017, p. 645). This issue of which rule applies in subsidiarity is a possible cause for conflicts and disputes, as two different methods of interpretation may be applied: one based on article 3(2) of the OECD Model Convention and another based on the VCLT. This also raises an alarm regarding a related issue concerning the legal status of the MLI explanatory statement under the VCLT’s general rules of interpretation. Should the explanatory statement conform to context within the meaning of Article 31(2) of the VCLT, then it is an important and binding element for interpreting the provisions of the MLI. Should it not belong to context, then it may be considered supplementary means of interpretation as provided for in Article 32 of the VCLT.
Since the MLI does not provide a solution for these possible conflicts between these various methods of treaty interpretation, it could give rise to disputes between states and ultimately to double taxation. Moreover, as the MLI does not provide for a supranational conflict resolution mechanism, including the establishment of an international judicial body or mandatory international arbitration, this could create fragmentation, through the creation of conflicting interpretations of the MLI provisions by various actors – domestic tax courts or arbitral tribunals.
Finally, interpretation and implementation conflicts and disputes regarding the provisions of the MLI can be resolved by a COP. The experience with a COP in Multilateral Environment Agreements (MEA) also raises questions as to the nature of COP decisions vis-à-vis international law. There is a fierce debate among scholars as to COP decisions being ‘hard’ or ‘soft’ law. Moreover, only a party to the MLI can initiate such a conference, so this is not an alternative open to taxpayers of the contracting parties.
Compatibility Clauses
Provisions of the MLI may conflict with provisions contained in the Covered Tax Agreements and in such cases this conflict is to be addressed using compatibility clauses.
In order to make the application of the MLI clearer and more transparent, where a provision replaces or modifies specific types of provision contained in a Covered Tax Agreement, the parties to the MLI shall make a notification listing all Covered Tax Agreements which contain such a provision. The effect of these required notifications will vary depending on the type of the compatibility clause.23
It is expected that the parties will use their best efforts to identify all the provisions that will subject to the scope of each compatibility clause. However, conflicts are possible either because parties to a Covered Tax Agreement disagree about whether a particular provision is within the scope of a compatibility clause or because they disagree about which provision is the relevant one (albeit agreeing that indeed there is a relevant provision for the purposes of the compatibility clauses). Such conflicts can be addressed through MAP provided for in the relevant Covered Tax Agreement of through COP.
Reservations and Notifications
As argued, reservations can materially impact the MLI scope of application. Moreover, notifications could also result in conflicts regarding their interpretation and implementation, principally in relation to the compatibility clauses referred to in the sub- chapter above. Article 28(1) establishes a list of allowed reservations and except for reservations pertaining to arbitration (Part VI of the MLI)24 these are the only reservations permitted. Once made, unless explicitly provided otherwise, reservations will modify MLI provisions to the same extent for both parties. Put differently, reservations will be applied symmetrically.
It is possible to withdraw a reservation or to replace it with a reservation more limited in scope. In order to withdraw or replace a reservation, the party must submit a notification to the Depositary. The purpose of allowing the replacement of a reservation is to also confer more flexibility to the MLI, thus increasing its scope of application and coverage throughout time. However, replacement cannot be used to place new restrictions or to enlarge the scope of existing reservations. This could induce parties to make as many and as broad reservations as possible, since reservations can be limited at a later time while the reverse route is not possible (BOSMAN, 2017).
Notification rules are extremely complex and detailed, but their overarching principle follows that of reservations. To regulate the notification process, Article 29 of the MLI provides for general notification rules and contains an exhaustive list of required notifications. Considering the complexity regarding notifications and their legal
According to the MLI, there are four of such clauses: “in place of”, “applies to or modify”, “in the absence of” and “’in place of’ or ‘in the absence of’”. For a through approach to this extremely sensitive topic, see (GOVIND; PISTONE, 2018).
Reservations regarding Part VI – Arbitration are covered by Article 28(2) instead.
consequences, there is the likelihood of possible notification mismatches25. Such conflicts can be addressed through MAP provided for in the relevant Covered Tax Agreement or through a COP.
Dispute Settlement Mechanism
BEPS contains a commitment to implement minimum standards, complemented by best practices, regarding the improvement of dispute resolution mechanisms. The objective is to lift the obstacles that prevent states from resolving treaty-related disputes in a timely, effective, and efficient manner. One of such minimum standards is the requirement that specific rules regarding MAP, as these rules are interpreted in the commentaries to the OECD Model Convention, are added to existing DTTs26. In order to facilitate the implementation and satisfaction of the above-mentioned minimum standard, Article 16 of the MLI allow parties to modify their Covered Tax Agreements to include MAP provisions.
Mutual Agreement Procedure – MAP
Article 16(1) through 16(3)27 adopts MAP as a means for settling conflicts and disputes arising out of the application of the Covered Tax Agreements. Accordingly, the competent authorities of the parties to such agreements shall endeavor to resolve, through MAP, double taxation situations. It is important to highlight here that states are required to seek a solution (endeavor to resolve), not to find one! This lack of an obligation to resolve disputes has created a stalemate in certain cases, thus resulting in double taxation not being relieved.
Pursuant to the commentaries to the OECD Model Convention, resolution by mutual agreement situations of double taxation situations is an integral part of the obligations assumed by a party which enters into a tax treaty and such obligation must be performed in good faith. Moreover, competent authorities are required to try to solve the case in a fair and objective manner in accordance with the terms of the relevant tax treaty and applicable principles of international law, namely the VCLT.
Although the legal status of MAP under domestic legal systems may vary28, the commentaries to the OECD Model Convention state that principles of international law on the interpretation of treaties, as contained in Articles 31 and 32 of the VCLT, allow domestic courts to rely on agreements reached through MAP. Furthermore, such
For instance, as stated in the sub-chapter covering compatibility clauses, a notification mismatch could occur if the parties to a Covered Tax Agreement disagree about whether a particular provision is within the scope of a compatibility clause, or if they have di fferent views as to which provision should be notified as relevant provision of a Covered Tax Agreement.
The article in the OECD Model Convention which deals with MAP is Article 25, paragraphs (1) through (5), however the agreed minimum standard only requires parties to adopt Articles 25(1) through (3).
These articles are a copy of Article 25(1) through 25(3) of the OECD Model Convention.
For a detailed analysis of the legal status of the OECD Commentaries under public international law, see (DOUMA; ENGELEN, 2008).
agreements represent objective evidence of the competent authorities’ mutual understanding regarding the meaning of certain provisions of a tax agreement and as such it must be taken into account for purposes of the interpretation of that tax treaty.
Through MAP, states have delegated to competent tax authorities broad and ample powers regarding the interpretation and implementation of the provisions of a tax treaty29. MAP is to be made available to taxpayers regardless of the existence of other eventual remedies – judicial or administrative – under the domestic systems. Put differently, MAP is independent from ordinary remedies available under national laws. However, most governments require that MAP and domestic remedies be not pursued simultaneously. In order to address this requirement, MAP is suspended while a proceeding before a domestic court or administrative tribunal is pending.
Whilst MAP has been perceived as a good solution to tax treaty disputes, some problems have been recognized and identified as well. It is an important mechanism to settle tax disputes between states, but it has experienced an increasing number of caseloads, with cases more than doubling between 2006 and 2015. Another relevant concern is the number of cases that remain unsolved or unresolved. To make matters worse, BEPS might further increase the number of new cases. Furthermore, MAP excludes another import feature of adjudication as it does not ensure that like cases will be treated alike, since MAP, in all its possible forms, is always processed under extreme secrecy.
Mandatory Arbitration
Mandatory binding arbitration has not been adopted. Indeed, this was one of the most contentious issue during the MLI’s negotiations as it implies that states will have to surrender their tax sovereignty to the arbitral tribunal and its arbitrators. However, this reliance on sovereignty grounds seems out of place not only as per the reasons already stated above, but also because a great portion of the doctrine considers that arbitration is simply an extension of MAP. Essentially it is a “spur to ensure resolution of cases by agreement between competent authorities” (PICCIOTTO, 2016).
Instead of being perceived as a third-party adjudication, arbitration should simply be seen as a tiebreaker rule between competent authorities. In other words, MAP is essentially an administrative procedure. Arbitration has been mainly structured as a continuation of the MAP in order to advance a predictable timetable for cases to be resolved (BROWN, 2015). In Brown’s own words “arbitration should be used to enhance mutual agreement procedure, not replace it” (Ibid, p. 8). However, mandatory arbitration, as a means of dispute resolution
Such powers include the authorization to resolve: (i) double taxation cases arising out of domestic laws, (ii) issues pertaining to the interpretation and application of the tax agreement itself, and (iii) to define a term not defined in the tax agreement or to clarify the definition of a defined term, in which case, the meaning attributed to a defined term through mutual agreement shall prevail over the meaning attributed to such term in the domestic laws of the parties
of tax disputes, has been met with resistance on various grounds. This resistance could be explained by different reasons.
First, the unwillingness of States to submit cases to an external decision since they understand this as an attack to their tax sovereignty.30 Second issue is that of equality. If states were to include arbitration provisions in their treaties, they would also have to grant similar procedure to citizens in purely domestic disputes, therefore implying that all domestic cases would have to become eligible for arbitration. This might be easily counter- argued with the argument that traditional non-discrimination principles, which exist in a multitude of legal orders, are commonly considered to require equal treatment only under same (or quite similar) circumstances and that is not the case of a pure domestic situation and one that entails two different jurisdictions. This being an overly complex and contentious topic, it would suffice to state that equality has been an argument that countries are particularly concerned with.
Due to such resistance, the OECD removed mandatory arbitration from the final report on BEPS. As mandatory arbitration has not been included as part of BEPS’ minimum standards, the MLI followed suit. Therefore, the MLI contains core arbitration provisions, but such provisions are not mandatory and may not be binding either 31.
As a final remark on mandatory arbitration, the MLI also opted for maintaining and guaranteeing the confidentiality of all arbitration proceedings by establishing in its Article 21 that only the members of the arbitration panel and a maximum of three staff of each party to the MLI may receive and have access to the information pertaining to the case. Secrecy may limit the ability of developing consistent and uniform interpretation of the MLI, for decisions will not be publicized, thus they may never form a set of jurisprudence to be used and adopted as authoritative interpretations by the parties. Under secrecy, there is no possibility of similar cases being treated similarly and that is a reason of concern and should be also explored in further detail.
Besides the above issues, all well discussed by scholars, another problem deserves special attention, the argument contended by some that MAP without an arbitration provision might violate fair trial standards32 33 under human rights laws. It can be said that most human rights conventions have been adopted without little attention being paid to tax
Another possible issue is the lack of engagement by developing countries. This lack of engagement could be explained by a higher level of sensitivity as arbitration might be perceived as a means to lose control over a country’s overall treaty policy and the pe rception among developing countries that foreign arbitrators predominantly support the interests of developed nations. See (ROCHA, 2017).
Another questionable provision in relation to mandatory binding arbitration, if chosen to be applied by the parties, refers to its non- binding nature. The arbitration award will not be implemented if the relevant competent authorities agree on a different resolution of all unresolved issues within three calendar months after the arbitration decision being delivered.
A right to a fair and public hearing within a reasonable time by an independent and impartial tribunal established by law.
Part of the doctrine defends that MAP does not violate fair trial standards because it is only a complement to domestic litigation procedures. Others advocate that even as an optional procedure, MAP should still fulfill fair trial requirements.
issues, but such human rights conventions have increasingly been applied in cases dealing with tax issues – both by the European Court of Human Rights (ECtHR) and by European domestic courts.
Notwithstanding the increase in its use, the ECtHR jurisprudence remains quite controversial.34 Be as it may, it is outside the scope of this paper to address the issue of human rights and tax35, its intention being simply to highlight that there is a growing concern that MAP, without a guarantee of an active participation of the taxpayer in the proceedings, may not fulfill fair trial requirements under international human rights laws. So, attention should be drawn to the issue of taxation and fair trail requirements as this could be of grower concern in years to come.
Conference of the Parties
Finally, regarding questions and disputes which may arise out of issues pertaining to the interpretation and implementation of the MLI itself, the parties to it may use a COP, which shall be convened pursuant to the Article 31, in order to take any decisions or exercising any appropriate and relevant functions that may be needed or required. After a close and attentive reading of the MLI, it becomes beyond question or doubt that its interpretation has been essentially left to States. No explicit role seems to be envisaged for any sort of third-party adjudication or a multilateral body to consider divergences in interpretation, let alone to issue authoritative opinions on them. This seems surprising, in view of the concerns expressed by many that BEPS reforms will lead to a period of great uncertainty.
Another relevant aspect relating to COP is its legal nature under international and domestic laws. In general, its legal nature is ambiguous, to say the least. International legal scholars have examined the legal status of COP decisions taken by treaty bodies and the vast majority of such scholars posit that COP decisions lack a binding nature under international law, unless the wording of the treaty explicitly says so. Additionally, to have a binding effect, COP decisions should express the parties’ intention to be bound by them.
Finally, it is also quite debatable if COP decisions would be viewed as binding under domestic tax systems. Whilst COP decisions could be used as guidance to the interpreter of the MLI, they will not be sufficient to create consistency for the purposes of reaching a harmonized interpretation of the MLI as each state might rely differently on the decisions reached by a COP. In that sense, it would have been advisable that the MLI had adopted a more global and uniform solution to deal with problems pertaining to issues of its own implementation and interpretation.
The leading case, as appointed by a good number of scholars, is the so-called Ferrazzini Case – (“Giorgio Ferrazzini v Italy, Merits, App
no 44759/98, 2001).
For a comprehensive analysis of the topic see (BAKER, 2001).
International Tax Adjudication
Notwithstanding the highly controversial nature of the use of international courts as an alternative36 way of settling international tax disputes, it is clear that the MLI has opted not to advance any discussion beyond MAP. The choice not to advance the discussions regarding international tax adjudication, although quite questionable, seems to be in close synchrony with the overall sentiment of the international community. Accordingly, there is little or no support for a dispute settlement mechanism based on a reasoned decision of an independent adjudicator. Even in cases where states have opted to make arbitration mandatory, the type of arbitration chosen, the so-called ‘baseball arbitration’37, has few of the features that can be expected of a reasoned and independent third-party adjudication.
According to Brown, the key feature of the LBO arbitration relates to the autonomy of the arbitration board which can resolve the case only by choosing the position of one of the competent authorities. The hope is that the competent authorities will be more reasonable in their negotiations, presumably because an extreme position is less likely to be adopted by the arbitration board. The true innovation of the US arbitration model, which has gained a lot of attention and seems to be the preferred model of arbitration adopted or to be adopted for tax purposes is that it forbids the arbitrators from providing a rationale for their decision. As said, the arbitrators will choose the best solution between the two put forward by the competent authorities (BROWN, 2015).
In spite of the interesting discussions surrounding LBO arbitration, what is clear for the purposes of this paper is that there is no intention by states to create any dispute settlement mechanism that would minimally resemble that of a true third-party independent adjudication. Even worse, any discussion of the appropriateness, feasibility, and desirability of an international court as an alternative means of dispute settlement mechanism remains almost restricted to academic discussions.38
In this sense, what is most striking is the fact that the idea of an international tax court has not been abandoned based on fears of fragmentation39 or others that have made the lives of international lawyers miserable for quite a long time. But, once again, due to the misconceived perception that tax is exceptional and different, and based on its historic
Given the predominance of the mutual agreement procedure as the only international means of settlement of tax treaty conflicts, it can be said that those newly emerging methods as well as the suggestions from legal doctrine, may be called “alternative” although adjudication by international courts and tribunals is of course hardly “alternative” from the perspective of general international law.
Baseball arbitration, technically called ‘Last Best Offer – LBO, contains no hearing or fact-finding phase (or discovery phase in a judicial setting) and prohibits arbitrators from independently deciding.
Only a few double taxation treaties provide for the possibility of an international court being the competent authority in a dispute settlement: the 1992 German-Swedish tax treaty refers to the International Court of Justice (ICJ) and the Austrian-German tax treaty which refer to the European Court of Justice (ECJ).
The phenomenon of “fragmentation” in international law, means the emergence of specialized and relatively autonomous spheres of social action and structure, which in turn leads to conflicts between rules or rule-systems, deviating institutional practices and, possibly, the loss of an overall perspective on the law.
isolationism, the discussion is still being put in terms of sovereignty! As we all know, sovereignty has been relativized enormously due to the human rights movement.
In summary, the issue is not the choice itself but the lack of a through debate regarding the issue of international adjudication of tax disputes. This is a pressing issue and is quite strange that it has not gained traction. The question that could probably be fairly raised is “who is afraid of international tax adjudication”?
CONCLUSION