The Multilateral Instrument: A New Array of Questions

The views and opinions expressed in this article are those of the authors only.

Lukas Kutilek*

The international tax regime as we know it today goes back to the beginning of the twentieth century. In 1923, the League of Nations reached a compromise on dividing the tax base between residence and source jurisdiction, which is usually called the Benefits Principle. Put simply, the Benefits Principle is the idea that active income should be taxed primarily at source and passive income should be taxed primarily at residency.[1] This principle is already embedded in the network of over 3,000 bilateral double tax treaties (“DTTs”). But the League of Nations also agreed on one other key principle of international tax regime. This second principle, sometimes called the Single Tax Principle, holds that although income should not be taxed twice, it should also not escape taxation altogether. The often-quoted language of the League of Nations states that:

“From the very outset, [the drafters of the model convention] realized the necessity of dealing with the questions of tax evasion and double taxation in co-ordination with each other. […] The most elementary and undisputed principles of fiscal justice, therefore, required that the experts should devise a scheme whereby all incomes would be taxed once and only once.”[2]

While it is mostly agreed that the Benefits Principle has been the basis for international taxation since 1923, the Single Tax Principle is more controversial.[3] Nevertheless, the focus is shifting from preventing double taxation to making sure that double non-taxation does not take place.[4] Whether one looks at the preamble to the 2016 US model DTT, the preamble to the OECD Multilateral Convention (“MLI”), the preambles to the hundreds of actual DTTs covered by the MLI’s minimum standards, the recitals of the EU Anti-Tax Avoidance Directive, the 2017 update to the OECD model DTT, or the proposed update to the UN model DTT, the Single Tax Principle seems to be omnipresent.

The OECD Secretary General Angle Gurria promised to “put an end to double non-taxation.”[5] The efforts resulted in the MLI signed by representatives of 71 jurisdictions on June 7, 2017. Is the MLI going to be successful in pursuing its noble goal to eliminate double non-taxation? Is much going to change for US taxpayers? To deal with these and other questions, the University of Michigan Law School hosted a conference labeled “Perspectives on the Multilateral Instrument,” which took place in Ann Arbor, Michigan on October 13, 2017. The conference was organized around five building blocks each consisting of a presentation and a following discussion.

Maikel Evers[6] presented on the conceptual approach of the MLI. Generally, the OECD needed to make sure it would efficiently implement anti-base erosion and profit shifting (“BEPS”) measures in a large network of bilateral treaties. What were the alternatives? Bilateral negotiations or a model protocol would take too much time and the result would be too uncertain. Therefore, the OECD decided to develop an instrument which would “sit on top of and modify bilateral tax agreements.” This approach is both innovative and allows for continuity; it modifies hundreds of treaties while maintaining their bilateral nature. The OECD also needed to strike a balance between having the most jurisdictions on board and implementing as many anti-BEPS measures as possible. This was done through a mechanism of minimum standards, opt-out provisions, and optional and alternative provisions.

Steven R. Ratner[7] commented on Evers with a short description of the bilateral investment treaty network and efforts to create a multilateral investment treaty. He concluded that the approach taken by the OECD with respect to the MLI could serve as a model to improve the investment treaty network as well.

Richard L. Reinhold[8] prepared a presentation on arguably the main MLI’s minimum standard provisions—the Principal Purpose Test (“PPT”). The PPT disallows treaty benefits if a transaction is carried out one of the principal purposes of which is to obtain such benefits, unless granting the benefits would be in accordance with the purpose of the treaty. It is a standard-based general anti-abuse rule (“GAAR”). Reinhold maintained that the PPT can be compared to an improvised explosive device in the hands of the tax authorities, which is overbroad and will catch taxpayers by surprise. He argued that a US-style limitation of benefits provision (“LOB”) would be a better tool to tackle anti-avoidance tactics.

In response, Reuven S. Avi-Yonah[9] expressed his view that the PPT is a better anti-avoidance tool than the LOB. First, the PPT disallows treaty benefits only if the purpose of the treaty is violated, and so concerns that all transactions will be barred are unjustified. Second, despite the LOB in the US model DTT, the actual LOB provisions are greatly more complicated and contain various loopholes. Third, the PPT’s uncertainty is the same as under domestic GAARs, which operate reasonably well and are not being abused by tax authorities.

Only practice can test the operation of the PPT. It is also likely that practice will vary among jurisdictions. Some commentators do not see any legal relevance in the PPT since teleological interpretation can yield the same result.[10] Others think the PPT will not operate differently than the “main purpose test” already included in the commentary to Article 1 of the OECD model DTT. It is true, however, that tax authorities will have at their disposal a very flexible instrument to attack aggressive tax positions.

Gianluca Mazzoni[11] presented his piece on one of the MLI’s specific anti-BEPS measures—Avoidance of Permanent Establishment Status. His presentation put the MLI in context with the recent tax avoidance cases of Apple and Google. Mazzoni argued that the MLI is the right framework to amend the permanent establishment concept (“PE”) to make sure that profits are taxed where value is created.[12] For example, Article 12 could be a successful tool in fighting commissionnaire arrangements as it lowers the dependent agent PE threshold to include enterprises, which habitually play the principal role leading to the conclusion of contracts. However, its effectiveness is undermined by the fact that 39 jurisdictions have opted out of Article 12.[13]

Pasquale Pistone[14] commented on Mazzoni’s presentation and agreed that reservations undermine the global effectiveness of the MLI. He also agreed that the PE concept does not work all that well anymore because of the way business is conducted nowadays. There is much evidence that businesses shape their strategies based on how their customers behave, and so information about the behavior is vital. Overall, Pistone concluded that in order to achieve a fair allocation of taxing rights around the world, we need to commit to revising the PE concept.

Pistone presented on cross-border tax dispute settlement and compared dispute resolution mechanisms in the MLI, the EU Arbitration Convention, and the proposed EU Arbitration Directive. The key improvement that the MLI introduces with respect to mutual agreement procedure (“MAP”) is a minimum standard provision allowing the taxpayer to make a request for a MAP to either tax authority. Nevertheless, one could argue that it is a missed opportunity that the MLI does not include other measures such as greater taxpayer participation, a timeframe for decisions, or the attention to taxpayers’ rights in general.[15] The MLI thus reinforces the traditional view that the MAP is carried out to define the boundaries of tax sovereignty, where the taxpayer is a mere object. This view, on the other hand, has been overcome by the new EU Arbitration Directive, which mentions taxpayer’s rights in the very first paragraph of its Article 1 and overall considers taxpayers as holders of rights as opposed as to mere objects. The MLI’s optional arbitration provision has been opted in by 26 countries, who all chose the so-called baseball arbitration, where the arbitration panel selects one proposed decision and does not issue a reasoned opinion. The arbitration decision also lacks any legal precedential value.

Kim Brooks[16] commented on Pistone’s presentation with four concerns. First, the baseball arbitration approach leaves out something important—the ability to build a baseline of knowledge from which others can learn. Second, the lack of publicity of precludes public scrutiny of the actions of the tax authorities. Third, the lack of arbitrators of choice may create a system where the same arbitrators will keep issuing the same decision in order to be reappointed. Fourth, taxpayers may still be more inclined to choose domestic dispute resolution mechanisms. These and other issues will have to be seen in practice. However, as Richard L. Reinhold indicated, the behavioral effect of baseball arbitration forces parties to price their case because they need to decide how much transaction costs they incur, which can in turn promote settlement. It is important to keep in mind that the implementation of anti-BEPS measures should not lead to unintended double-taxation or uncertainty for compliant taxpayers, which is why a functioning dispute resolution mechanism is an integral part of the MLI.[17]

Reuven S. Avi-Yonah concluded the conference with his presentation on the MLI and the international tax regime. He started by posing a provocative question: Why do we need double tax treaties at all when double taxation is mostly eliminated unilaterally by domestic law? The reason seems to be the prevention of double non-taxation, i.e. supplementing the Benefits Principle by the Single Tax Principle. After all, one must first define and secure the tax base before dividing it among jurisdictions. This was the reason to adopt the MLI and from this perspective, it is a great achievement. Avi-Yonah also noted that the MLI is important from the US perspective because it represents a shift in the global international tax regime. Victor Thuronyi[18] responded with a brief criticism of the OECD’s choice to give the MLI the form of a non-textual amendment instead of a plain textual amendment. In short, non-textual amendments generally have the same effect as textual amendments, but are not the cleanest method of legislative drafting.

The authors and participants contributed to a very lively and up-to-date discussion. Many questions were answered, and other were raised. But only an operation in practice will show whether the MLI is a revolutionary achievement or a blind alley in the explicit implementation of the Single Tax Principle.


* Lukas Kutilek is a JD candidate at the University of Michigan Law School and holds a JD equivalent, summa cum laude, from the Charles University in Prague.

[1] For a brief overview of both theoretical and pragmatic justification for the Benefits Principle see Reuven S. Avi-Yonah, International Taxation of Electronic Commerce, 52 Tax L. Rev. 507, 556 (1997).

[2] Report Presented by the Committee of Technical Experts on Double Taxation and Tax Evasion, League of Nations Doc. G.216.M.85 II (1927).

[3] For an example of the on-going debate, see H. David Rosenbloom, International Tax Arbitrage and the “International Tax System” 53 Tax L Rev. 137 (2000) and Reuven S. Avi-Yonah, Commentary, 53 Tax L Rev. 167 (2000). See also Michael Lang, Introduction to the Law of Double Taxation Conventions 45 (2010) (“It is occasionally concluded that the purse of DTCs also includes the avoidance of double non-taxation. In my view, however, this is incorrect in such general terms.”).

[4] Hugh J. Ault, Some Reflections on the OECD and the Sources of International Tax Principles, 70 Tax Notes Int’l 1195, 1195 (June 17, 2013).

[5] Webcast: OECD presents outputs of OECD/G20 BEPS Project for discussion at G20 Finance Ministers’ Meeting OECD (Oct. 5, 2015), http://www.oecd.org/ctp/oecd-presents-outputs-of-oecd-g20-beps-project-for-discussion-at-g20-finance-ministers-meeting.htm.

[6] Maikel Evers leads the work on the MLI at the OECD Centre for Tax Policy and Administration.

[7] Steven R. Ratner is the Bruno Simma Collegiate Professor of Law at the University of Michigan Law School

[8] Richard L. Reinhold is Senior Counsel in Willkie Farr & Gallagher LLP and Adjunct Professor of Law at the New York University School of Law.

[9] Reuven S. Avi-Yonah is the Irwin I. Cohn Professor of Law and director of the International Tax LLM Program at the University of Michigan Law School.

[10] Michael Lang, BEPS Action 6: Introducing an Antiabuse Rule in Tax Treaties, 74 Tax Notes Int’l 655, 663 (May 19, 2014).

[11] Gianluca Mazzoni is an SJD Candidate at the University of Michigan Law School, as well as an LL.M graduate from the same university. He is also a guest contributor to the Oxford Business Law Blog.

[12] To tax profits where the value is created is one of the leading principles of the BEPS project, included also in the preamble to the MLI.

[13] 39 countries opted out of Article 12, Artificial Avoidance of Permanent Establishment Status through Commissionaire Arrangements and Similar Strategies.

[14] Pasquale Pistone is Academic Chairman of the International Bureau of Fiscal Documentation, a Jean Monnet ad Personam Chair in European Tax Law and Policy at Vienna University of Economics and Business, and Associate Professor of Tax Law at the University of Salerno.

[15] Pasquale Pistone and Phillip Baker previously described the lack of attention to taxpayer’s right in MAP quite deplorable. P. Pistone & P. Baker, General Report, in The Practical Protection of Taxpayers’ Fundamental Rights 65-66 (IFA Cahiers Vol. 100B 2015).

[16] Kim Brooks is Professor of Law at Schulich School of Law, Dalhousie University.

[17] See Action 14: 2015 Final Report in OECD/G20 Base Erosion and Profit Shifting Project: Making Dispute Resolution Mechanisms More Effective 9 (OECD ed., Oct. 5, 2015), http://www.oecd.org/ctp/making-dispute-resolution-mechanisms-more-effective-action-14-2015-final-report-9789264241633-en.htm.

[18] Victor Thuronyi is a former Lead Counsel at the International Monetary Fund.