MANDATORY BINDING DISPUTE RESOLUTION IN THE BASE EROSION AND PROFIT SHIFTING (BEPS) TWO PILLAR SOLUTION

Abstract

Binding taxpayer-initiated international dispute resolution has traditionally played a minor role in the international tax system. Despite being long pursued by corporate interests and increasingly accepted by developed countries, international tax arbitration has remained less developed and less respectful of private interests than investor–State arbitration. The binding multilateral dispute settlement endorsed by over 130 countries as part of the Organisation for Economic Co-operation and Development's Two Pillar Solution to issues raised under Action 1 of the Base Erosion and Profit Shifting (BEPS) project marks a change and is noteworthy at a time when some States are reconsidering their consent to the international adjudication of trade and investment disputes. The design of international dispute settlement in the Two Pillar Solution, and the focus on the protection of multinationals from juridical double taxation, displays little appreciation of the experience with dispute settlement in international trade and investment over the past two decades.

I. INTRODUCTION

Global tax governance is characterised by intergovernmentalism with specks of multilateralism and is supported by a network of shared norms and expectations.¹ Putting aside the European Union (EU), multilateral treaties have had only a marginal impact on the setting of the rules for cross-border income taxation.² The cross-border income tax regime derives its structure from bilateral double tax treaties, and from non-binding standards for cross-border taxation of income and capital agreed in the Organisation for Economic Co-operation and Development (OECD) and, to a lesser extent, the United Nations (UN).³ Double tax treaties allocate the rights to tax between countries and collectively constrain the approach that an individual jurisdiction can take to cross-border income tax issues. The purpose of these treaties is to minimise the risk of juridical double taxation and tax discrimination of non-residents. Over 3,000 bilateral double tax treaties are in force around the world.⁴ They follow either the OECD or the UN Model Tax Convention,⁵ which are themselves very similar and reflect the conceptual structure of the first model treaty drafted by a League of Nations Committee of Technical Experts in 1927.⁶ The approach to dispute settlement in double tax treaties has traditionally been distinct from trade or investment treaties, shunning both binding State–State and binding State–taxpayer dispute settlement. Many double tax treaties have a Mutual Agreement Procedure (MAP). The procedure is based on Article 25 of either the OECD Model Tax Convention or the UN Model Tax Convention and allows taxpayers to challenge actions of States participating in a double tax agreement when these actions result in taxation that is not in accordance with the agreement.⁷

Over the last 15 years, developed countries have moved to accept mandatory arbitration in double tax treaties. An option of binding arbitration was included in the MAP article of the OECD Model Tax Convention in 2008 following long-term corporate advocacy.⁸ In 2013, the OECD and Group of Twenty (G20) launched the Base Erosion and Profit Shifting (BEPS) project.⁹ BEPS refers to tax planning strategies that exploit gaps and mismatches in tax rules to avoid paying income tax—usually by shifting profits to low- or no-tax jurisdictions where there is often little or no real economic activity.¹⁰ In 2015, over 60 countries agreed to 15 Actions under the BEPS project. BEPS Action 14 sought to enhance the scope and effectiveness of dispute settlement mechanisms in bilateral double tax treaties, including through taxpayer access to mandatory binding arbitration.¹¹

In 2016, the OECD and G20 members established an Inclusive Framework, a structure which allows interested countries and jurisdictions to work on an equal footing with the OECD and G20 members in the next phase of the BEPS project. This structure currently includes 141 countries.¹² In 2017 many of these countries signed a Multilateral Instrument to Prevent BEPS which, among other things, gave effect to rules developed under BEPS Action 14.¹³ Signatories were free to opt out of some of the Instrument's provisions. As a result, only about 30 countries, mostly developed countries and low- or no-tax jurisdictions, committed to mandatory binding arbitration for certain tax disputes.

In October 2021, corporate interests obtained the inclusion of mandatory binding arbitration as part of a core component of the so-called Two Pillar Solution to the income tax challenges of digitalisation.¹⁴ Despite traditionally strong resistance to taxpayer-initiated international tax arbitration outside of the OECD, especially from developing countries, and concerns raised by tax and political science scholars,¹⁵ 138 members of the Inclusive Framework agreed to an unprecedented role for binding dispute settlement within the international income tax regime without much fanfare.¹⁶ This sudden turn of events warrants attention.

International dispute settlement will be available unevenly across different components of the Two Pillar Solution. Most significantly, Pillar One of the Two Pillar Solution will create a new taxing right, labelled ‘Amount A’, for the jurisdictions in which certain very large multinationals sell their products and derive non-routine profits.¹⁷ The proposed multilateral convention that will establish the right to collect Amount A will permit multinationals to elect to have matters related to Amount A resolved in a mandatory and binding manner if State actions create a risk of juridical double taxation.¹⁸ An arbitration decision relating to Amount A would bind all jurisdictions that are party to the multilateral convention establishing the right to collect Amount A.

The embrace of taxpayer-initiated, mandatory and multilaterally binding arbitration by the members of the Inclusive Framework is out of step with developments in international economic law.¹⁹ The triumphalist rhetoric of international economic lawyers of two decades ago has disappeared, because of the sovereignty and legitimacy concerns of developed and developing countries created by international dispute settlement. The United States, for example, caused the World Trade Organization (WTO) dispute settlement system to collapse in 2019. Many States are rejecting or seeking to limit binding investor–State dispute settlement. The EU and other States are seeking to develop an appellate mechanism for investor–State dispute settlement because of the perceived shortcomings of ad hoc international arbitration. Tax arbitration procedures have attracted increased attention in the tax literature, but with only limited engagement with the challenges that dispute settlement is facing in other areas of international economic law.²⁰ So, it is both surprising and unsurprising that the recent embrace of binding mandatory dispute settlement in international taxation has occurred without any apparent reference to the challenges that dispute settlement has faced in trade and investment treaties. This article seeks to understand the likely future of international tax arbitration under the Two Pillar Solution in light of the existing role of arbitration in double tax treaties and the experience of other international economic law dispute settlement processes. For different reasons, the experiences under the international trade and investment regimes suggest that Two Pillar Solution dispute settlement processes and the ongoing Two Pillar Solution policy development and implementation will generate growing concerns with legitimacy and sovereignty. A rosy future is not assured. Greater investment in ensuring the legitimacy of dispute settlement processes is required.

The article proceeds as follows. Sections II and III examine how MAP arbitration and mandatory binding dispute resolution under the Two Pillar Solution is likely to operate. Section IV asks what lessons the international income tax regime may learn from the recent collapse of the WTO Appellate Body. Section V argues, despite the unique features of international tax arbitration, that many of the key concerns expressed about WTO dispute settlement and investor–State dispute settlement (ISDS) are applicable to binding arbitration under the Two Pillar Solution. Section VI concludes.

II. MAP ARBITRATION

A. Arbitration in International Taxation

Recognising that income from cross-border activities might be taxed in more than one jurisdiction, the model tax treaties provide that the country where income is generated (the source State) should have priority over the country of the taxpayer's residence (the residence State), and that active (business) income should be taxed at source, while passive (investment) income should be taxed on a residence basis. Much tax planning and cross-border income tax avoidance by multinationals takes advantage of the structure of the international income tax system, often by shifting profits to low- or no-tax jurisdictions where there is often little or no real economic activity. The efforts of one State to collect corporate income tax, close loopholes and respond to new forms of business organisation and tax planning may be challenged by a taxpayer or another State as inconsistent with a double tax treaty. Contentious cross-border tax rules and application of tax treaties are mostly litigated by taxpayers in national courts.²¹ While cross-border tax matters are not infrequently the subject of inter-State consultations, the model tax conventions do not contain a standalone State–State dispute settlement mechanism. Double tax treaties of developed States, however, often have MAP provisions for resolving difficulties that arise out of the treaty.²²

Like investor–State dispute settlement, the MAP is an outgrowth of the process of diplomatic protection in general international law whereby a State asserts a claim against another State because one of its nationals has been treated in a manner that is in violation of international law.²³ Under the MAP, on the request of the taxpayer, the competent authorities of the Contracting States engage with each other and endeavour to resolve disputes that arise from the way that one or both contracting States have interpreted or applied the tax treaty.²⁴ The issues which may be subject to the MAP include transfer pricing adjustments, attribution of profits to permanent establishments, determination of residence of individuals and companies, deduction of withholding taxes, application of tax treaty anti-abuse provisions and the application of domestic anti-avoidance provisions. The competent authorities are required to endeavour to resolve the case by mutual agreement. If agreement is reached, it must be implemented by the contracting States.²⁵ The competent authorities may also resolve by mutual agreement any difficulties or doubts arising as to the interpretation or application of the tax treaty and eliminate double taxation in cases not provided under the treaty.²⁶

The number of MAP cases commenced each year has been growing since 2016.²⁷ In 2020, 2,508 cases were started, and 1,725 cases were closed.²⁸ Of the cases started and closed, 1,178 and 667 were transfer pricing cases. In 2018, just over 75 per cent of MAP cases were resolved, that is, the relevant competent authorities were able to reach agreement on the matter. The process is less effective in resolving the ‘big’ cases.²⁹ MAP cases are very concentrated, with 25 jurisdictions accounting for 95 per cent of the cases started in 2020.³⁰

Consistent with its origins in diplomatic protection, the MAP requires the competent authorities to act in good faith but does not require them to reach an agreement or to disclose their exchanges to the taxpayer.³¹ As noted in the Introduction, over the past decade and a half, tax law, following the broader trend in international law, started giving more attention to individual remedies under international rules.³² While most do not, the MAP provisions in some treaties now permit the taxpayer to submit issues that the competent authorities cannot resolve by mutual agreement to binding arbitration.³³ This was occurring at a time when some States were becoming increasingly wary of and withdrawing their consent to international dispute settlement in other areas of economic law, which reflects the general detachment of international tax law from other areas of international law.

Before the 2017 Multilateral Instrument to Prevent BEPS was signed,³⁴ a little more than 200 double tax treaties included an arbitration procedure, and even then often arbitration was neither mandatory nor binding and the treaties did not specify the procedure for the arbitration.³⁵ Under Article 25(5) of the OECD Model Tax Convention, only claims of taxation not consistent with the tax treaty which were not resolved within two years of the submission of all relevant information to the competent authorities may be submitted to arbitration.³⁶ Precedence is, however, given to domestic court processes. The taxpayer is unable to initiate arbitration if a decision on the issues to be submitted to arbitration has already been rendered by a court of administrative tribunal of either State.³⁷

The MAP provisions in Article 25 have been criticised on many grounds, including the limited access to them, uncertainty, lack of transparency, long timeframe and no guarantee of resolution of the dispute.³⁸ Some steps to address these concerns were taken under BEPS Action 14 and the Multilateral Instrument to Prevent BEPS. Countries that join the Inclusive Framework must commit to implementing the 15 Action Items identified by the OECD to combat BEPS and meet the four Minimum Standards. The fourth Minimum Standard relates to treaty disputes and arbitration. Changes to the MAP processes brought about under the Inclusive Framework, and the signing of the Multilateral Instrument to Prevent BEPS, have improved both the effectiveness of MAPs and access to arbitration.³⁹ The Instrument, however, contains only optional provisions for the settlement of taxation disputes by arbitration, because of objections to mandatory arbitration from many jurisdictions.

The mandatory arbitration clause in the Multilateral Instrument to Prevent BEPS is a modified and more detailed version of the clause in Article 25(5) of the OECD Model Tax Convention. The Instrument did not formally amend double tax treaties but modified some of the rights and obligations of the parties to those treaties because it was a more recent treaty under Article 30(3) of the Vienna Convention on the Law of Treaties.⁴⁰ Article 30 of the Multilateral Instrument to Prevent BEPS expressly preserves the rights of the parties to a double tax treaty to amend that treaty. The bilateral obligations created by tax treaties and the Multilateral Instrument to Prevent BEPS differentiate MAP arbitration from a multilateral dispute settlement mechanism. Under double tax treaty provisions like Article 25, disputes are resolved through application of the relevant double tax treaty rather than domestic law. This will probably increase the importance of the Explanatory Statement adopted at the time of the Multilateral Instrument to Prevent BEPS and any OECD Commentary on arbitration proceedings.⁴¹

As of December 2022, 100 jurisdictions have signed the Multilateral Instrument to Prevent BEPS.⁴² If all 100 ratified the Instrument and made no reservations, up to 1,700 tax treaties could be modified. However, the Instrument only modifies double tax treaties in force between parties to the Instrument which have been listed by both parties as covered by the Instrument.⁴³ Modification is also subject to parties’ reservations. So far, around 650 bilateral double tax treaties have been modified.⁴⁴ The United States has, however, not signed the Multilateral Instrument to Prevent BEPS.⁴⁵ Nonetheless, the Multilateral Instrument to Prevent BEPS entered into force on 1 July 2018.

So far, the Instrument has had a limited observable impact on tax arbitration. As noted above, the mandatory binding arbitration provisions of the Instrument (Articles 18 to 26) only apply if both jurisdictions to a dispute have expressly decided that their bilateral double tax treaty is a covered tax agreement. Only 30 jurisdictions have opted in. Furthermore, few (if any) of these jurisdictions are likely to be connected by double tax treaties to all of the other 29 jurisdictions that are opting in.⁴⁶ The Multilateral Instrument to Prevent BEPS also permits the jurisdictions that opt in to make reservations and therefore limit the scope of the mandatory arbitration clause,⁴⁷ and over half of jurisdictions opting in have chosen to do so.⁴⁸

As with Article 25 of the OECD Model Tax Convention, once an arbitration decision has been delivered, the contracting States are expected to implement it through the MAP.⁴⁹ However, an arbitration decision does not need to be implemented if the contracting States agree on a different solution.⁵⁰ The taxpayer also has the option to reject the arbitration decision or litigate in a national court.⁵¹ Arbitration decisions can also be declared invalid by national courts.⁵²

In 2016, an International Fiscal Association report stated that apart from a couple of cases, most States had no experience in tax arbitration, observing that ‘it is evident that taxpayers cannot rely on this mechanism in the way [it] is implemented today’.⁵³ A couple of years later the number of arbitrations appears to have grown. Mann cites data from Europe, where in 2018 out of a total of 942 pending MAP cases there were 47 arbitrations, and in 2019 out of 778 pending MAP cases there were 42 arbitrations.⁵⁴ If slightly over 5 per cent of MAP cases are going to arbitration that is not insignificant. The prospect of mandatory arbitration will encourage the competent authorities to reach an agreement that is in line with the double tax treaty. On the other hand, greater familiarity with the arbitration process is probably leading to a greater number of cases.⁵⁵ The use of arbitration in international tax matters appears to be growing, albeit from a low base.

B. Limitations of MAP Arbitration

Developing countries remained wary of the risks associated with the arbitration of tax disputes when the Multilateral Instrument to Prevent BEPS was developed.⁵⁶ They had multiple concerns, including loss of sovereignty, their experience with investor–State dispute settlement, substantial litigation and revenue loss costs, and limited experience in or capacity to arbitrate. As a consequence, binding mandatory arbitration became optional in the BEPS project. Article 25(5) modified under the Multilateral Instrument to Prevent BEPS has a number of limitations as a dispute settlement mechanism and did not meet all of the expectations of corporate taxpayers.

Only cases where taxes have actually been charged can be submitted to arbitration under the Multilateral Instrument to Prevent BEPS.⁵⁷ This condition would appear to preclude advisory decisions and may reduce the opportunities of corporate taxpayers to use MAP arbitration strategically. Nonetheless, MAP arbitration is still seen as primarily a benefit for taxpayers and its inclusion in the BEPS Action 14 Minimum Standard was supported by the business community. Taxpayers have little to lose from arbitration and cannot be really said to be bound by the process or the outcome. The Multilateral Instrument to Prevent BEPS does not contain a fork-in-the-road clause typical of investment treaties. Taxpayers are generally free to exit the MAP process and pursue remedies under domestic law at any time,⁵⁸ but cannot pursue MAP and domestic remedies simultaneously.⁵⁹ Even after the competent authorities have negotiated a resolution or the matter has been arbitrated, the taxpayer is not bound to accept the outcome.⁶⁰

On the other hand, the competent authorities can also agree on a longer or shorter period than the usual two-year period before MAP arbitration can be commenced.⁶¹ If any of the competent authorities fail to appoint a member of the arbitration panel in the manner and in the timeframe specified in Article 19(2) of the Instrument, a member is appointed by the highest-ranking official of the OECD Centre for Tax Policy. However, without a default set of rules for arbitration, delays or the withholding of agreement on the rules will prevent arbitration proceeding. The Instrument rules therefore cannot be seen as providing the same consent to international arbitration normally present in investment treaties.⁶²

Two types of arbitration are foreseen under the Instrument rules: baseball type arbitration and legal opinion arbitration. Legal opinion arbitration anticipates a reasoned decision as might be expected in an international trade dispute settlement or investor–State arbitration. Under baseball arbitration, each party puts forward its preferred solution, and the arbitration panel must choose between the two solutions.⁶³ Baseball-type arbitration aims to narrow the differences between the parties and discourage extreme positions because they are likely to be rejected by the arbitrators. A jurisdiction may make reservations as to the type of arbitration in which it is prepared to engage. Jurisdictions party to a dispute may therefore have incompatible preferences, in which case arbitration cannot go forward until the parties agree on the format of the arbitration.⁶⁴ The process for baseball-type arbitration may be speedier than legal opinion arbitration. However, decisions will have no precedential value and lack transparency. While the legal opinion approach formerly does not create a precedent, the existence of legal opinions helps to clarify legal rules and advance legal certainty. Their existence is likely to affect the future actions of tax administrations and arbitrators even if they are subject to broad confidentiality obligations.⁶⁵

MAP decisions are generally not made public, which may be explained by the process being rooted in diplomatic protection, the nature of baseball arbitration, taxpayer privacy and commercial secrecy concerns, and a wish to avoid the creation of precedents.⁶⁶ The lack of transparency and legal certainty under the MAP is often criticised.⁶⁷ The OECD admitted that publishing reasoned arbitral decisions ‘would lend additional transparency to the process’ and ‘having the material in the public domain could influence the course of other cases so as to avoid subsequent disputes and lead to a more uniform approach to the same issue’.⁶⁸ Unless MAP arbitration decisions are reasoned and published, they will not influence the interpretation or application of the tax treaty by national courts.

In summary, prior to the Two Pillar Solution being agreed, arbitration was available under only a small fraction of the over 3,000 bilateral tax treaties, and, even when it was available, consent to be bound by the arbitration by either the contracting States or the taxpayer was highly qualified.

III. BINDING DISPUTE SETTLEMENT IN THE TWO PILLAR SOLUTION

The Two Pillar Solution, reached after lengthy and at time difficult negotiations, was the global response to the income tax challenges arising from the digitalisation of the economy.⁶⁹ A full account of the negotiations or the outcome is beyond the scope of this article, but some background is necessary. Despite the modest role played so far in international taxation, international tax dispute resolution was a key element of the negotiations leading to the Two Pillar Solution. This section examines the scope for the use of international dispute settlement procedures to resolve issues arising from the Two Pillar Solution. Different aspects of the Two Pillar Solution were matched with different international dispute settlement responses, ranging from none to the creation of a radical multilateral process for the new taxing right ‘Amount A’.

The income tax challenges arising from the digitalisation of the economy are central to BEPS Action 1.⁷⁰ Cross-border online commerce was thriving and the companies operating the international digital platforms had grown rapidly and were making enormous profits but were paying little or no income tax in most of the jurisdictions where their customers were located. The ability to structure a business so that little or no corporate income tax is paid in many jurisdictions where customers are located, and from which significant profits are derived, may have been taken to a new level by the digital platforms, but is not unique to them. The digital platforms nonetheless provided the political imperative for governments to take some action.

The United States as the home State to most of the largest digital platforms was not keen on changes to the century-old framework for the international allocation of rights to tax income that would increase the tax bases of States importing the digital services and goods sold online. The combination of the political imperative for action in market States and the unwillingness of the United States and others to contemplate a change that was outside the conceptual structure of the existing international income tax regime led to policy innovation. A key innovation was the introduction or threat of the introduction of digital services taxes on some suppliers of digital services. Such taxes were not necessarily inconsistent with the WTO and essentially fall outside double taxation treaties, but the United States was firmly opposed and reacted strongly.⁷¹

The Two Pillar Solution is a compromise suggested by the OECD's Secretariat in response to the demands of the United States. In the version accepted by Inclusive Framework members in October 2021, Pillar One aims to ease tensions between the home jurisdictions and market jurisdictions of large multinationals, essentially, by agreeing to a new taxing right for market jurisdictions known as Amount A in exchange for a prohibition on the use of digital services taxes and similar taxes. The introduction of Amount A will require a level of multilateral coordination and cooperation not seen previously in taxation.

Pillar Two aims to address a different set of concerns arising from corporate income tax competition. It seeks to incentivise all countries to apply a minimum level of corporate income taxation for multinational enterprises, but without tax rates being set under a binding treaty. The key rules that were agreed are modified versions of tax measures unilaterally introduced by the United States.

A. Pillar One

The Pillar One rules seek to reallocate, for the purposes of corporate income tax, some of the profits of very large and very profitable multinationals and multinationals engaged in specific intra-group transactions.⁷² The two main legal elements of Pillar One are an agreement on a new taxing right for market jurisdictions (Amount A) with a binding dispute resolution procedure, which would be implemented through a multilateral convention and domestic legislation;⁷³ and a domestic transfer pricing rule for market jurisdictions known as Amount B.⁷⁴ Only some host countries will be able to collect both Amount A and Amount B from qualified multinationals. A draft of the proposed multilateral convention for Amount A is yet to be released. Of the publicly available information, the 225-page October 2020 Pillar One Blueprint sets out the overall design of the Amount A rules and institutions. Subsequently, the OECD has released for public consultation draft model rules for some aspects of Amount A, but not yet for international dispute settlement.⁷⁵

The proposed multilateral convention for Amount A will contain a broad array of institutional and procedural rules to make the substantive rules workable, as well as a prohibition on the use of digital services taxes and similar measures. On 20 December 2022, the OECD released for public consultation draft provisions for the removal of all existing digital services taxes and other relevant similar measures.⁷⁶ The multilateral convention will create mutual obligations on taxation for all jurisdictions that are party to the convention, whether or not they are presently connected through a bilateral double tax treaty. While only a relatively small fraction of pairs of jurisdictions have committed to binding arbitration in their double tax treaties, under the multilateral convention for Amount A multinationals will be able to elect to have issues related to Amount A resolved in a mandatory and binding manner if they are at risk of juridical double taxation.⁷⁷ The focus in the OECD documentation is on providing certainty to multinationals—on request—rather than assisting tax administrations challenging multinationals’ self-assessment or the validation of the self-assessment by the lead tax administration.⁷⁸ The proposed Amount A dispute settlement system gives prominence to the norm against juridical double taxation, which protects corporate interests in relation to the Amount A and creates a precedent for the future.

Aside from Pillar One, virtually all tax disputes are handled in bilateral processes, and sometimes a series of bilateral processes, rather than through a multilateral process.⁷⁹ Corporate taxpayers, however, can seek multilateral dispute prevention and/or resolution for Amount A. The design of the Amount A allocation rules, which involves an assessment of income of the entire multinational group, probably necessitates some sort of multilateral process, because any dispute between two jurisdictions over Amount A will probably affect the taxation of Amount A in multiple jurisdictions.⁸⁰ Baseball arbitration will not work for a dispute over Amount A or similar tax bases.⁸¹ The Pillar One Blueprint proposes that a panel of a representative sample of tax administrations from jurisdictions where the multinational operates will conduct a substantive review of the multinational's self-assessment. Tax administrations from all jurisdictions affected by the allocation of Amount A will have an opportunity to object to the panel's decision. The dispute resolution process is then envisaged to operate as follows:⁸²

• • If the review panel cannot reach agreement or cannot accommodate objections from other tax administrations, questions will be referred to a determination panel that is required to reach a decision, which is binding on all tax administrations.

• • Rules for dispute prevention and resolution could be embedded in the same instrument that introduces rules for the taxation of Amount A, ensuring that the new taxing right is linked to the availability of the new and enhanced tax certainty regime.

These commitments will go far beyond what is offered under the 2017 Multilateral Instrument to Prevent BEPS. Indeed, the process is intended to address some of the limitations of MAP arbitration, including its voluntary and bilateral nature. The process envisages a very significant delegation of authority in relation to Amount A to a determination panel. The arbitration will take the final determination of the rules governing Amount A out of the hands of individual States. The Pillar One Blueprint, however, proposes that the determination panel result will not bind the multinational that triggered the process.⁸³ Moreover, the tax administrations would be bound even if the multinational withdraws from the process.⁸⁴ This is an unusual arrangement. There are no risks to a multinational commencing the dispute settlement process because they can walk away at any point. The multinational will be able to pursue other remedies against a State for not complying with the treaty or domestic law, possibly including when the State is complying with the decision of the determination panel. The existence of multiple remedies will strengthen the negotiating position of the multinational with tax administrations and contribute to multi-jurisdictional litigation strategies.

The ability of taxpayers to reject the outcomes of MAP arbitration has been justified on the basis that the taxpayers have a constitutional or human right to access the courts, and the taxpayer is not party to the arbitration.⁸⁵ The first reason flows from the second. The consent to arbitration of participating jurisdictions and the rights of the multinational under the Pillar One Blueprint proposal are now closer to investor–State arbitration than the traditional MAP arbitration.

An elective binding dispute resolution mechanism is also being considered for eligible developing economies.⁸⁶ Again, something beyond the enhancements of tax arbitration offered in the Multilateral Instrument to Prevent BEPS must be contemplated. The eligibility of developing countries will be reviewed regularly. Once found ineligible by a review, jurisdictions will remain ineligible in all subsequent years.⁸⁷ The prospect for this mechanism is unclear. Unless the rules for developing country-initiated arbitration are part of the multilateral convention establishing the Amount A taxing rights, the jurisdictions that are home to most of the multinationals will have little incentive to implement the agreed dispute settlement mechanism.

The other main component of Pillar One, Amount B, is a new transfer pricing rule for the allocation of business profits related to transactions between members of a multinational group. The rule will apply to ‘in-country baseline marketing and distribution activities’ of subsidiaries or permanent establishments of a multinational group in the market jurisdiction and will standardise the remuneration for these activities.⁸⁸ The objective is to increase the profits that will be subject to corporate income taxation in market jurisdictions. The introduction of the Amount B rule by a market jurisdiction will require that jurisdiction to secure agreements to amend its double tax treaties. Any disputes related to the application of Amount B in a way which creates the risk of double taxation under transfer pricing rules may also be subject to mandatory binding dispute resolution if a relevant double tax treaty allows.⁸⁹ Amount B will therefore be subject to different dispute settlement rules than Amount A. In fairness, Amount B is likely to be subject to fewer disputes than Amount A because Amount B is founded on a fixed rate of return on baseline marketing and distribution activities.⁹⁰

The publication of well-reasoned arbitration decisions would seem to be important to reducing uncertainty over how the provisions will be implemented.⁹¹ In this regard, the past practice in relation to MAP arbitration is not reassuring. In any event, the multilateral process envisaged for Amount A dispute resolution may require creative thinking to address issues associated with the sharing of commercially sensitive information while protecting the rights of all jurisdictions to object.

B. Global Anti-Base Erosion Rules

Implementation of Pillar Two of the Two Pillar Solution would create three new domestic tax rules (together known as the Global anti-Base Erosion (GloBE) rules) and a model treaty rule (the Subject to Tax Rule) for double tax treaties.⁹² The three GloBE rules are the Qualified Domestic Minimum Top Up Tax on the excess profits of a constituent entity of a multinational located in the jurisdiction; the Income Inclusion Rule, which imposes a top-up tax on a parent entity in respect of the low taxed income of a constituent entity, which may be in or outside the jurisdiction; and the Undertaxed Payment Rule, which denies deductions or requires an equivalent adjustment to the extent that the low tax income of a constituent entity is not subject to tax under the Income Inclusion Rule.⁹³ The rules are complex, but the core idea is that where a source jurisdiction imposes a corporate income tax of less than 15 per cent on an entity of a multinational enterprise, another jurisdiction that hosts a parent or related entity of the same multinational will be able to impose a top-up tax that equals the difference between the actual level of taxation in the source jurisdiction and 15 per cent. There are, of course, exceptions and further rules setting out the top-up taxing rights and priorities of those rights for multinational enterprises with entities in multiple jurisdictions. The GloBE rules address more than the headline corporate income tax rate and will limit the ability of countries to provide tax holidays to attract investment, which are common in developing countries. Some workarounds will be possible, but difficult where the source jurisdiction is a low- or no-tax jurisdiction where no real economic activity occurs, ie, a tax haven. The GloBE rules are designed, it is claimed, to dampen corporate income tax competition heavily even if only the major capital exporting jurisdictions implement the full suite of GloBE rules.

The GloBE rules are intended to be domestic and voluntary but, according to the OECD, ‘will have the status of a common approach’, which means that members of the Inclusive Framework on BEPS ‘are not required to adopt the GloBE rules, but, if they choose to do so, they will implement and administer the rules in a way that is consistent with the outcomes provided for under Pillar Two, including in light of model rules and guidance agreed to by the IF [Inclusive Framework]’; and ‘accept the application of the GloBE rules applied by other IF members including agreement as to rule order and the application of any agreed safe harbours’.⁹⁴ These are significant commitments. For a voluntary arrangement, it is highly prescriptive, except for the relationship between GloBE and the United States’ Global Intangible Low-Taxed Income (known as GILTI) regime,⁹⁵ which is left open in the OECD documents. The two taxes have similar aims, and the United States tax provided the inspiration for the GloBE rules, but they are different in important ways.⁹⁶

A range of tax disputes may arise under the GloBE rules, notwithstanding the claim that the rules ‘have been designed in a way to minimise the scope for disputes concerning their application across multiple jurisdictions’.⁹⁷ First, some commentators have argued that GloBE may itself breach many double tax treaties.⁹⁸ Similar but contested claims have been made in relation to GILTI and Base Erosion Anti-Abuse Tax (BEAT) taxes in the United States.⁹⁹ In October 2021, the OECD stated that at the latest by the end of 2022 an implementation framework will be developed that facilitates the coordinated implementation of the GloBE rules.¹⁰⁰ At a minimum, a switchover rule, which is a mechanism that would enable home jurisdictions to amend or override double tax treaty obligations under which they have committed to exempt profits attributable to foreign permanent establishments, is needed. Without the switchover rule, the GloBE rules could not treat exempt branches the same as foreign subsidiaries to prevent low-taxed income of a foreign branch being combined with high-taxed income in the home jurisdiction. If this view is correct, the fact the GloBE rules are domestic and not treaty based means that they might be challenged under a double tax treaty, if they are inconsistent with the treaty. Secondly, the Qualified Domestic Minimum Top Up Tax or the top-up tax under the Undertaxed Payment Rule may be applied to a multinational in a manner inconsistent with the Two Pillar Solution. Thirdly, both top-up taxes may be applied by more than one State to the same multinational. The second and third possibilities appear to be the motivation for an agreement providing for binding mandatory arbitration, because multinationals are at risk of juridical double taxation.

The proposed mechanism for implementation of the GloBE rules does not contemplate a multilateral convention. MAP arbitration is only possible where the relevant States are already connected by a double tax treaty and, even then, arbitration is usually not available to resolve disputes. This suggests that mandatory binding dispute resolution will not generally be available for Pillar Two issues. In any event, the availability of MAP arbitration along the lines of Article 25 of the OECD Model Tax Convention may not be able to resolve some types of dispute. The possibility that top-up taxes are imposed by more than one jurisdiction leading to juridical double taxation may require multiple jurisdictions to be party to a dispute to allocate taxing rights consistently, unless all these jurisdictions amend domestic legislation with provisions that create a MAP obligation for them.¹⁰¹ It would be possible for States to commit to arbitrate certain tax disputes in national legislation as part of the ‘common approach’.

If binding dispute settlement was available, the issue of the relationship between the United States’ GILTI regime and the GloBE rules might be impossible to avoid. Assuming that the GILTI regime is not in conformance of the common approach agreed in the GloBE rules, the United States practice might be challenged. This possibility is unlikely to be acceptable to the United States. Assuming, however, that a dispute settlement challenge by a multinational to the imposition of an additional tax under the GILTI regime was possible, a jurisdiction might wish to seek a negotiated resolution with the United States to avoid an arbitration decision. A finding that the GILTI regime is inconsistent with a tax treaty or other international instrument may simply force the United States to demand the amendment to the international agreement or other action. Many States would not wish to see their relationship with the United States be put at risk.

C. Subject to Tax Rule

The Subject to Tax Rule will allow source jurisdictions to tax certain related-party payments (eg, interest; royalties; franchise fees; insurance premiums; guarantee, brokerage or financing fees; rent payments for movable property; and consideration for the supply of marketing, procurement, agency or other intermediary services between two companies which are part of the same multinational group) that are taxed in the jurisdiction where the taxpayer received the payment (which will usually be the taxpayer's residence jurisdiction) at a nominal corporate income tax rate below 9 per cent.¹⁰² The Subject to Tax Rule would only apply between developed and developing countries.

The implementation of the Subject to Tax Rule will require the amendment of a jurisdiction's double tax treaties because the rule would affect treaty commitments made in relation to taxation of passive income. The Inclusive Framework members are yet to agree on whether the Subject to Tax Rule will be implemented by means of a model treaty provision which would need to be incorporated in each double tax treaty by its amendment or through incorporation of the Subject to Tax Rule in a multilateral instrument that will modify rights under double tax treaties.¹⁰³ Either option may well lead to partial or delayed implementation of the Two Pillar Solution. Once incorporated into the relevant tax treaties, the Subject to Tax Rule would be subject to the MAP, including arbitration if available.¹⁰⁴ Otherwise disputes relating to the income tax in the recipient or source jurisdiction are subject to domestic processes. The OECD commentary on the Subject to Tax Rule will probably be influential for interpretation in both domestic processes and the resolution of disputes through a MAP.

Without a comprehensive multilateral instrument, the legal rights under Pillar Two will be asymmetrical. A foreign multinational operating in a developed country objecting to the imposition of a Subject to Tax Rule tax would be able to invoke the MAP in a double tax treaty. A developing country objecting to the failure of a developed country to agree to the amendment of a double tax treaty would not have similar rights and would not have a right to countermeasures under general international law unless the commitment of Inclusive Framework members to the Subject to Tax Rule is ultimately contained in a binding legal instrument. Where there is no double tax treaty, no dispute settlement provisions will be applicable unless they are contained in a multilateral instrument. The United States has been prepared to agree to bilateral tax arbitration similar to Article 25 of the OECD Model Tax Convention but would be likely to baulk at a multilateral instrument with mandatory arbitration provisions. Unlike Amount A, dispute settlement is not or less essential here to protect US multinationals from juridical double taxation.

D. Uncertain Future

The future of the Two Pillar Solution is not settled. The Pillar Two report observes that ‘It may be possible to include the GloBE provision in the new multilateral instrument considered under Pillar One, which could also have the benefit of setting out the interaction between Pillar One and Pillar Two.’¹⁰⁵ This now seems unlikely. Even so, it will only concern jurisdictions that see a benefit in Amount A relative to the costs and benefits of digital services taxes and similar taxes. A less positive view of the benefits of Pillar One for many States is possible after they have considered the legislation and administrative steps required to collect their share of Amount A.

The Two Pillar Solution will be implemented through an array of binding and non-binding instruments. The multilateral convention that is intended to give effect to Amount A and the prohibition on digital services taxes will presumably not enter into force until it has achieved a specified critical mass of ratifications. The number will be set to reassure the parties that enough jurisdictions are committed so that convention provides net benefits to its members. When a jurisdiction ratifies the convention, that not only creates international obligations and rights but also affects the interests of both parties and non-parties, including in relation to their own ratification decisions. Jurisdictions that contemplate ratifying later in time will consider the identity and reservations of other jurisdictions that have ratified. Moreover, if a decision to ratify is made, reservations may be crafted in response to the ratifications and reservations of other parties. Putting aside issues that may arise from the possibility of reservations, some jurisdictions are likely to hold off ratifying the Pillar One multilateral convention quickly. Dispute settlement may be one of the key issues.

Most jurisdictions will have little incentive to ratify the proposed multilateral convention for Pillar One's Amount A until it is clear that the United States will also ratify. They will not wish to be caught with an obligation to implement the convention when the home jurisdictions of the in-scope multinationals are not parties to the convention. Implementation of Amount A rules without a multilateral convention may result in a jurisdiction breaching, inter alia, the equivalent of Articles 5, 7 and 9 of the OECD Model Tax Convention that are in many double tax treaties. If the United States wishes to act in a manner that is consistent with its existing double tax treaties, it will need to amend its laws. It is not clear how easy it would be to take amendments to multiple double tax treaties and introduce the multilateral convention for Amount A through the United States Senate. The politics of corporate income tax and the aversion to a loss of sovereignty and binding dispute settlement will be significant obstacles.

The situation with Pillar Two is different. On 22 December 2022, the Council of the EU issued the Minimum Tax Directive which requires Member States to incorporate the Pillar Two rules into domestic law by 31 December 2023. EU members with no more than 12 ultimate parent entities of in-scope multinational enterprises, however, can delay introduction of the Directive until 31 December 2029.¹⁰⁶ As noted above, the United States already has rules that are similar to but not compliant with the GloBE rules. There is no consensus on which non-EU jurisdictions will have a strong interest in the implementation of Pillar Two.¹⁰⁷ Much depends on how jurisdictions make decisions under conditions of uncertainty. Arguably, some jurisdictions will have little incentive to ratify a multilateral instrument implementing Pillar Two quickly unless the United States links that to its ratification of Pillar One. If that view is correct, the failure of the United States to ratify the multilateral convention for Pillar's One Amount A may cause the previously widespread support for Pillar Two to unravel. The major capital-exporting countries will still have the legitimation of a unilaterally implemented rule that protects their tax base and countries as destination for investment.

The value of Pillar Two to jurisdictions other than the United States may, in part, depend on the relationship between the GloBE and the similar tax provisions already in place in the United States (GILTI and BEAT). If the ‘common approach’ has no treaty basis, inconsistent approaches to the GloBE rules and double taxation are the likely consequences. A treaty basis for GloBE would, however, put certain tax exemptions at risk in many jurisdictions. The risk of politically significant tax exemptions being examined in a MAP arbitration is likely to be politically problematic for some countries, including the United States. If this analysis is correct, unilateralism and treaty overrides seem the likely outcome.

The final binding and non-binding instruments intended to give effect to the Two Pillar Solution have not been finalised, accepted and implemented by the members of the Inclusive Framework. The future of the package and its elements is not certain. Inclusive Framework members have decisions to make, including on the role of dispute settlement in any new international tax rules. In making those decisions, members should understand the political limits of adjudication of disputes in international economic law and explicitly address the need for legitimacy and public accountability. These issues are taken up in Sections IV and V below.

IV. POLITICAL VIABILITY OF INTERNATIONAL DISPUTE SETTLEMET

Establishing and maintaining binding international dispute settlement are not easy. The experience in the WTO suggests that even if mandatory binding arbitration of international tax disputes is agreed, sustaining that system over the decades will face challenges. The challenges are likely to grow as the number of cases and the ‘size’ of the cases increase. The Dispute Settlement Understanding (DSU) was part of the grand bargain that produced the WTO.¹⁰⁸ The more power-oriented approach to dispute settlement under the General Agreement on Tariffs and Trade (GATT) is often contrasted with the conscious move to the more rule-oriented WTO processes.¹⁰⁹ Under the DSU, members no longer had the ability to delay or block the formation of a panel, agreement on terms of reference, appointment of panellists, adoption of panel recommendations, or authorisation of retaliation. The DSU created a right of appeal to the Appellate Body, which was often described as the ‘crown jewel’ of the WTO. Within a decade of the establishment of the WTO, the Appellate Body emerged as one among the most widely used and strongest international courts and enjoyed broad support from the WTO members.¹¹⁰ Despite popular perceptions of the Appellate Body as a world trade court, the intention was never to create a strict rule of WTO law. Political oversight and, in theory, control were essential elements.

In 2018, taking advantage of the consensus rule for decision-making, the Trump administration decided to block all appointments to the Appellate Body.¹¹¹ By December 2019, the Appellate Body was left with just one member. The last three members of the Appellate Body agreed to continue working on the three appeals for which a hearing had been held after their terms expired.¹¹² Work on all other appeals was indefinitely suspended. Without an Appellate Body decision, the disputes remain unresolved, and the challenged measures remained in place. The Biden administration has continued to block the appointment process.

Since the demise of the Appellate Body, almost all of the disputes have been appealed ‘into the void’, preventing the adoption of the panel report by the Dispute Settlement Body. A further consequence has been the dramatic drop in the number of WTO disputes that have been commenced. In 2018 there were 38 requests for consultation and in 2019 there were 20. By contrast, in 2020 there were five and in 2021 there were nine.¹¹³ In 2021, ten panels and one compliance panel were established, and seven reports relating to nine disputes were circulated, but eight disputes were appealed. As of 31 December 2022, 24 disputes were pending before the Appellate Body.¹¹⁴

The reasons for the collapse of the WTO dispute settlement are salient to the growing role of binding dispute settlement in the international tax regime. The United States has voiced a number of complaints about the Appellate Body and WTO dispute settlement system but has not (yet) identified what it thinks needs to be done to fix the system. The Appellate Body was said not to have followed its own rules, including ignoring mandatory deadlines for completion of reports; allowed Appellate Body members to continue to sit after the expiry of their terms; reviewed questions of fact and not just law; treated Appellate Body reports as binding precedents for panels; opined on matters not necessary to resolve a dispute; and sought to determine matters that are within the authority of WTO members.¹¹⁵ The substantive concerns of the United States focused on trade remedy cases that went against the United States, in particular those prohibiting ‘zeroing’ and the defining ‘public bodies’. While some of these concerns were shared by other members,¹¹⁶ the scholarly consensus is that the United States’ complaints were overstated if not outright wrong.¹¹⁷ The United States has declined, and continues to decline as of mid-2022, to engage with proposals from other members that would be the restoration of the Appellate Body.¹¹⁸ The recent statements of the United States in the WTO indicate that it is not ready to agree to a restoration of binding dispute settlement, notwithstanding United States participation in informal consultations about WTO reform.¹¹⁹ In the meantime, the United States has commenced and defended panel proceedings and appealed some but not all panel reports into the void.

The conduct of the United States in the WTO is consistent with the long-standing aversion to binding third-party dispute settlement in the United States. The United States has not ratified treaties because, at least in part, of opposition to dispute settlement provisions and refused to submit to the general jurisdiction of the International Court of Justice, Inter-American Court of Human Rights, International Tribunal for the Law of the Sea, or International Criminal Court.¹²⁰ Such conduct is consistent with the preference, identified by international relations scholars, of hegemonic powers for diplomatic rather than legal dispute settlement, where they can deploy their power to greatest advantage. Only a rather unique set of circumstances encouraged members of the United States Senate to accept that the downside of the WTO dispute settlements system was outweighed by the rest of the WTO in 1994.¹²¹ The strong support of corporate business in the United States and the wishful expectation that arbitration will largely only constrain other jurisdictions might suggest that tax arbitration could receive a pass. A more realistic outcome is that arbitration might constrain United States tax legislation, and therefore will be rejected by the United States Senate, which, it has been argued above, might put pressure on the Two Pillar Solution.

A focal point of the United States government's discontent with the DSU was the outcome of several disputes with China.¹²² The outcomes were the result of fairly literal interpretations of the WTO Agreement on Subsidies and Countervailing Measures. Discontent with the decisions therefore is not entirely consistently with the United States’ concern about judicial activism. The source of the discontent with the decision lay ultimately in the economic and geopolitical challenge today perceived to be posed by China, but not anticipated when China acceded to the WTO in 1995.¹²³ The same concerns may over time afflict dispute settlement under the Two Pillar Solution if Pillar One is implemented. As economic and geostrategic environments change, the major powers may find the Two Pillar Solution rules, or certain interpretations of the rules produced by arbitrators, less convenient, in which case continued adherence to the regime or individual decisions is unlikely.

Other countries have maintained a greater commitment to dispute settlement in international economic law. The EU's response to the collapse of the Appellate Body was to establish a Multi-Party Interim Appellate Arbitration Arrangement,¹²⁴ which could provide effective appellate review to the signatories on the basis of Article 25 of the DSU and guarantee the adoption of a panel report by the Dispute Settlement Body. Australia, Brazil, Canada, China and Japan are among the 48 signatories. India and most African and Asian WTO members are not a party. The parties in several disputes have already agreed to use the Multi-Party Interim Appellate Arbitration Arrangement. The EU also takes compliance with WTO panel decisions seriously.¹²⁵ The EU's commitment to multilateral processes is not, however, unconditional. The EU has indicated its willingness to resort to unilateral countermeasures under general principles of international law,¹²⁶ and use dispute settlement provisions in its preferential trade agreements.¹²⁷ Nonetheless, the collapse of the Appellate Body is widely seen as likely to bring about a longer-term shift away from a rule-based trading system.¹²⁸

The selective attachment of major powers to international law does not go unnoticed. Some developing countries have also been prepared to sabotage regional courts that have ruled against them.¹²⁹ Only a third of WTO members chose to join the Multi-Party Interim Appellate Arbitration Arrangement. Small countries will be circumspect about likely outcomes of agreements to arbitrate a cross-border tax issue, if large States refuse to submit to binding WTO dispute settlement or do not implement panel recommendations, especially in the light of the underdeveloped mechanisms for the enforcement of MAP tax arbitration decisions.¹³⁰ Non-compliance by large States might facilitate small States doing likewise. Under the Two Pillar Solution smaller countries may then have the ability not to comply with arbitration decisions that go against them.

The Appellate Body crisis was fuelled by the inability of the WTO members to clarify and develop the WTO agreements through negotiation. Being left to interpret ambiguous provisions, the Appellate Body was criticised for being overly textual. Textualism provided a device for the Appellate Body to insulate itself from the political controversies within and surrounding the WTO.¹³¹ The features of MAP and Pillar One's Amount A arbitration discussed above might temporarily shield decisions from scrutiny and decision-makers from accountability. On the other hand, experience with dispute settlement in other contexts shows that these same features will generate legitimacy challenges (see Section V below). It is unclear whether the Two Pillar Solution dispute settlement processes will be able to walk what appears to be the narrow path¹³² required to maintain political and professional community support and doing what it does in a manner that comports with the basic requirements of the rule of law.

Further challenges may come from different quarters. The Two Pillar Solution affects not only individual taxpayers but mandates tax policies that affect multiple public interests. The anti-globalisation movement has attacked the WTO for not being adequately concerned with the environment, health or human rights.¹³³ One basis for the criticism is that these values are required to be considered because members have binding obligations under other areas of international law. The principle of the harmonious interpretation of different rules of international law could play a greater role in the international tax regime than the WTO.¹³⁴ Sustainable development and the right to development¹³⁵ are directly affected by international tax rules, and the response to BEPS issues, and arguably have a role to play in the interpretation of international tax instruments, even though these issues have been traditionally side-lined in OECD tax policy-making processes.¹³⁶ If non-tax considerations are invoked, commercial interests may be concerned. With greater prominence will come greater politicisation of tax arbitration.

One lesson from the WTO is that the establishment of dispute settlement regimes is a political act, and the resilience of any regime cannot be guaranteed. Tensions created by institutional design weaknesses, changing geo-economic circumstances, and litigation losses may grow over time. The WTO DSU was never expected to enforce a strict rule of trade law, but even the inbuilt flexibility proved insufficient, and the future of the binding dispute settlement in the WTO has been put into question by the United States. These considerations suggest that the signs for binding dispute settlement under the Two Pillar Solution are not auspicious. The rights conferred on multinational taxpayers differentiates the situation from the WTO. However, the Inclusive Framework members have failed to heed the lessons learnt from investor–State dispute settlement in the design of their mechanisms.

V. SOME COMPARATIVE INSTITUTIONAL DESIGN LESSONS

Global governance has seen the rise in novel forms of public authority and the demise of a sharp national–international divide.¹³⁷ Dispute settlement under the Two Pillar Solution fits that broader trend. In this regard, the international rax regime bears some resemblance to the international investment protection regime. Both are primarily structured by several thousand bilateral agreements that share many common provisions, and protect the private economic interests in capital-exporting countries.¹³⁸ The Two Pillar Solution and the international tax regime in general, however, embody a different balance between bilateral and multilateral rights compared to the investment and trade regimes. International tax has been on the move from a bilateral towards a multilateral order.¹³⁹ A long-standing controversy is whether the WTO agreements create a bundle of bilateral obligations or whether the agreements pursue a common purpose.¹⁴⁰ The situation in the tax regime is clearer. Some of the international obligations to be created under the Two Pillar Solution appear to be multilateral (eg, rules relating to Amount A), while others are bilateral (eg, Subject to Tax Rule) even where the rules have their origin in a multilateral instrument. Notwithstanding the limitations on MAP arbitration discussed in Section IIB and structural differences, the experience under the international investment regime speaks most directly to the design choices being made for dispute settlement under the Two Pillar Solution.¹⁴¹

The number of investor–State arbitration cases rose significantly after the mid-1990s (see Section IIA). Now, international tax arbitration case numbers look set to increase, if not follow a similar trajectory, which has cost and political implications for many countries. The total investor–State dispute settlement (ISDS) case count had reached over 1,100 by the end of 2020. Claims of over US$100 million are common.¹⁴² To date, 124 countries and one economic grouping are known to have been respondents to one or more ISDS claims.¹⁴³ The amount at stake in tax arbitration could in due course surpass that claimed in ISDS. The costs of bringing and defending a WTO dispute are in the millions and the costs of a typical ISDS case could be five times higher.¹⁴⁴ The costs of tax arbitration are also substantial. All members of the Inclusive Framework do not have the same financial or technical capacity to bring and defend Two Pillar Solution-related tax disputes against well-resourced multinationals.

The legal capacity of a member has been shown to matter more than its power in WTO disputes.¹⁴⁵ Empirical work also shows that poorer countries do worse in ISDS cases than richer countries.¹⁴⁶ The same concerns have been raised by developing countries in the past as one of the reasons not to embrace tax arbitration.¹⁴⁷ Unsurprisingly, the BRICS (Brazil, Russia, India, China and South Africa) have actively sought to increase their legal capacity in the WTO over the past two decades, neutralising to some extent an advantage long held by large developed members.¹⁴⁸ The loss of advantage in dispute settlement is broadly correlated with the decline in the acceptance of the WTO dispute settlement outcomes by the United States. An increase in the number of international tax arbitrations will mean a significant increase in the amount of tax revenue affected, which in turn means that international tax arbitrations will no longer be able to fly below the public and political radar. Any perceived unevenness of the playing field in tax arbitration will attract attention and may have longer-term systemic effects on international tax cooperation.

The Pillar One Blueprint proposes that experts could provide technical assistance to developing countries or those countries with resource constraints.¹⁴⁹ The design and funding of any technical assistance programme will determine its effect. A general comment without specifics does not provide much reassurance. Something akin to the Advisory Centre on WTO Law¹⁵⁰ would be a necessary piece of a programme of assistance if developing countries are to engage in tax arbitration effectively.¹⁵¹ If developing countries were likely to be as well-resourced as the multinationals arbitrating tax disputes, support for international arbitration among multinationals might not even be so firm.

Unlike WTO disputes, MAP and Amount A arbitration can be initiated by private actors,¹⁵² and they share certain similarities with investment disputes. International tax arbitration will have an impact on the past and future tax liabilities of companies and the revenue collected by tax jurisdictions. As already observed, at a time when opportunities for taxpayer-initiated international tax arbitration appear to be expanding, governments in many developed and developing countries are re-examining and reducing their commitments under investment treaties, because of concerns about ISDS.¹⁵³ The concern is not just with restrictions on host State policies under investment agreements, but also with ISDS as a process and therefore directly relevant to tax arbitration. The ISDS provisions in investment agreements have been vigorously criticised by civil society groups, politicians and scholars throughout the world.¹⁵⁴ ISDS is attacked for the absence of the level of accountability expected in democratic societies, the unequal access to legal representation, and a failure to meet rule-of-law principles. Many of the features which render arbitration more attractive than litigation in domestic courts—the possibility for discrete settlement or a formal ruling without publicity, the avoidance of appellate review, and the prospect of selecting adjudicators—are the same features that are seen as raising rule-of-law concerns. Accordingly, a number of recent investment agreements have sought to limit ISDS or excluded it entirely.¹⁵⁵ The UN Conference on Trade and Development (UNCTAD) has been working to assist States to modernise older-generation investment treaties. In parallel, the United States and other developed countries have also elaborated their model investment agreements. The United States and its business community, however, remain supportive of ISDS in principle. The same is true of China and Japan. Perhaps indicating that if MAP and Amount A arbitration are unlikely to constrain, in practice, the tax policies of major capital-exporting countries and at the same time assist their multinationals, an imperfect system of tax arbitration may still receive political support.

Many of the features of the ISDS process that have been challenged as illegitimate are being replicated in the international tax regime. The legitimacy of a regime exerts a ‘compliance pull’ on those normatively addressed by the regime.¹⁵⁶ Decisions of international tribunals and courts that are perceived to be illegitimate are less likely to be complied with. In the investment regime, the Convention on the Settlement of Investment Disputes¹⁵⁷ and New York Convention¹⁵⁸ have mitigated risks of non-compliance with awards. Similar arrangements are not in place for international tax arbitrations, and the ease and reliability with which tax arbitration decisions can be enforced in many jurisdictions are uncertain and have not so far been addressed in the Two Pillar Solution. Consequentially, if tax arbitration is not perceived to be illegitimate, the impact on compliance may be greater than in the case of ISDS.

The obligation to implement the outcome of the arbitration is clearer for MAP arbitration than it is for the WTO. In the WTO, a panel or the Appellate Body normally will only make a recommendation to bring the offending measure into compliance.¹⁵⁹ That recommendation needs to be adopted by the Dispute Settlement Body, which is a committee composed of all WTO members, before any requirement to implement arises.¹⁶⁰ The MAP article in double tax treaties results in an arbitrated agreement between the home and host State. Parties to an arbitration have a duty to implement the decision, unless the decision is rejected by the taxpayer, but no wider process of enforcement cooperation exists.¹⁶¹ The publication of trade and investment arbitration decisions adds to the pressure on governments to comply with the decision. As discussed below, tax arbitration suffers from a very low level of transparency. The design of Two Pillar Solution dispute settlement will not aid the self-enforcement of its arbitration decisions and may incentivise States to explore the weaknesses of the arrangements for enforcement of arbitration awards. Multinational taxpayers may need to call on their home States to assist, which will politicise individual cases.

States have been concerned about inconsistent and expansive holdings by ISDS arbitration panels, which in turn deters and/or raises the costs of implementing public policies.¹⁶² Specific process-related concerns have included the selection of arbitrators, lack of transparency, no third-party participation, lack of an appeal process, and prioritising market rationality over fundamental values of the international community. Many of the process-related concerns with ISDS would seem to be applicable to MAP and Amount A arbitration or any similar process agreed. The social class, education, career and expertise of arbitrators, for example, have been shown to shape ISDS.¹⁶³ Similar concerns have been raised about tax arbitration, in particular the perception that arbitrators will be dismissive of or hostile to the concerns of developing countries.¹⁶⁴

The International Centre for the Settlement of Investment Disputes (ICSID), the most important forum for ISDS arbitration, has negotiated new arbitration rules to address some of the process-related criticisms.¹⁶⁵ Discussion of broader and more significant procedural changes are underway under the auspices of the UN Commission on International Trade Law (UNCITRAL), but are unlikely to bear fruit for several years.¹⁶⁶ The European Commission has proposed replacing the ad hoc arbitration panels with a more centralised investment court system, inspired in a large part by the WTO Appellate Body.¹⁶⁷ The discussion of permanent international courts or tribunals and appellate mechanisms has been a prominent part of the UNCITRAL's work. Critics fear an investment court will be dominated by a governmental ethos, with little guarantee of more impartial or competent arbitrators.¹⁶⁸ The OECD documentation on the Two Pillar Solution did not discuss these ongoing negotiations.

While some issues, such as an absence of an appeal process for ISDS, remain contentious, it is unclear why the Inclusive Framework members, or even the more active OECD countries in the development of the Two Pillar Solution, felt limited pressure to address some of the features of ISDS that are broadly accepted as problematic. One possible explanation may be the domination of representatives from tax administrations in the negotiations. If this is correct, broadening country representation in the Inclusive Framework beyond national tax administrations may help raise awareness of any process-related deficiencies.

As discussed above, taxpayers are generally free to exit the MAP process and pursue remedies under domestic law at any time. There is no fork-in-the-road provision in MAP arbitration treaty articles, despite being common in investment agreements. Even after the competent authorities have negotiated a resolution of the matter, the taxpayer is not bound to accept the outcome.¹⁶⁹ On the other hand, jurisdictions can prevent a matter being resolved by arbitration by agreeing to a solution before or after the arbitration has commenced.¹⁷⁰ Tax arbitration is in some respects worse, from the point of view of the public interest, than ISDS because the taxpayer can renounce the outcome of the arbitration. On the other hand, the involvement of the competent authorities and their ability to exit the process through mutual agreement may provide some constraint on the taxpayer.

Jurisdictions will have less freedom in relation to Amount A, but the multinational will still be free to reject the arbitration decision. The one-sided commitment is open to abuse by taxpayers, and it will affect the perceived legitimacy of the process. If Amount A needs a multilateral dispute settlement process, a multinational enterprise should be bound by the outcome of a process that it initiates.

Tax arbitration will be less transparent than investment or trade disputes. The MAP arbitration procedure, it will surely be objected, was intended to resolve individual tax cases simply, rather than developing general interpretations of a treaty.¹⁷¹ A baseball-type arbitration by its nature cannot be expected to provide transparency or precedent. The legal opinion process, however, also arguably lacks the transparency and guarantees of rationality expected of a decision that has significant consequences for the public. While strong protection for confidential information is necessary,¹⁷² the failure to discuss transparency and accountability in Amount A arbitration is particularly surprising given the practice and trend in investment arbitration. The inability to share analyses and outcomes of cases works against the consistent application of rules and learning by domestic authorities. It may also have an impact on arbitrator selection, favouring those with experience and creating a privileged club of arbitrators.

Even with rules clarifying that arbitration decisions do not create binding precedents, a tension will inevitably arise between the pursuit of transparency and legitimacy, on the one hand, and non-precedential decisions, on the other hand. Published decisions will affect future decisions. Notwithstanding that Article 3(2) of the WTO DSU provides that the DSU intended to ‘clarify provisions’ of the WTO agreements in accordance with customary rules of interpretation of public international law, and ‘Recommendations and rulings of the [Dispute Settlement Body] cannot add to or diminish the rights and obligations provided in the covered agreements’, the United States’ attack on the Appellate Body was based on a concern that the Appellate Body was in fact affecting rights and obligations. The issues to which it is proposed that binding dispute settlement will be available under the Two Pillar Solution reflects the economic interests of the large, developed countries and their multinationals. Without arbitration decisions being made public, large States will have a greater ability to reject decisions unilaterally or pressure smaller States to agree to another outcome. The asymmetrical availability of binding dispute settlement does not appear to be a good recipe for long-term stability. Together with adequate legal capacity, transparency may help make interaction between smaller developing countries, and multinationals and their home States more rule oriented.

Rights to unilaterally terminate an investment agreement are usually limited, and, when terminated, existing investments will still benefit from protection for many years.¹⁷³ Nonetheless, recent years have seen an upsurge in terminations and, more frequently renegotiation of treaties, which can produce more immediate effects.¹⁷⁴ It is not clear how a jurisdiction may withdraw from part or all of the Two Pillar Solution, in particular the Amount A multilateral convention. Buyer's remorse in relation to the Two Pillar Solution might arise if States were unable to assess accurately the consequences of the package at the time of signature or ratification, but over time the costs and benefits became clearer as treaty provisions were applied and better understood.¹⁷⁵ States that have been involved in ISDS, in particular those that are challenged, have been the States most likely to seek renegotiation or termination.¹⁷⁶ In other words, many States are not good at assessing the costs and benefits of treaties ex ante. The same might be expected in the Two Pillar Solution. The renegotiations of investment treaties, interestingly, focused more on substantive policy space, with more modest changes to ISDS processes.¹⁷⁷ It may, of course, be that States most unhappy with ISDS seek to terminate rather than renegotiate.

Once a jurisdiction has ratified the proposed Amount A multilateral convention modifying its rights and obligations under double tax treaties, the practical freedom to reject some of the Two Pillar Solution rules may be limited. Changing the rules would require the renegotiation of several, possibly dozens of double tax treaties. If the counterparties are unwilling to negotiate, a jurisdiction may be forced to choose between renunciation of the tax treaty or living with the relevant Two Pillar Solution rules. The opportunity for treaty override will depend on constitutional structure of each jurisdiction.¹⁷⁸ Dualist (or effectively dualist) countries like Australia, Canada, Germany, the United Kingdom and the United States appear to have more freedom to move. More consideration would need to be given to the amendment and exit from treaty arrangements.

The consensus-based decision-making in the WTO has resulted in almost total gridlock, which in turn resulted in too much being asked of the WTO's dispute settlement system. The gridlock is in part due to the shift in power away from developed countries toward emerging economies in the world economy and the WTO. The large emerging economies have learned how to utilise WTO process to their advantage as they have grown in economic importance.¹⁷⁹ The more decentralised investment treaty regime, by contrast, has allowed a degree of learning and evolution of treaty rules.¹⁸⁰ The Two Pillar Solution, combined with the structural features of the existing international income tax regime, seeks to create a regime that is more centralised than the investment regime. It is also a regime that the developed countries expect to be able to dominate and supplement with further implementing arrangements without fear of being held up because of a lack of global consensus. It is not clear that the OECD countries will be able to dominate the tax regime in the longer term. The major emerging countries are building their capacity in international tax law and policy. Many developing countries also have a preference for tax to be negotiated under the auspices of the UN rather than the OECD.

The Two Pillar Solution contemplates many new international instruments. As noted at the outset, non-binding standards have played an important role in the international tax regime. The processes by which Two Pillar Solution guidelines and commentary are being developed have been and are likely in the immediate future to be dominated by a fairly small group of States.¹⁸¹ As is common throughout global governance, many States will not send representatives, and others will send generalists often from the local embassy and sometimes with authority simply to take notes rather than to speak.¹⁸² Lacking expertise, some governments have sent private arbitration lawyers to represent them in the ongoing investment agreement reform negotiations in the UNCITRAL.¹⁸³ Experts that spend their career working in tax are likely to have both the requisite expertise to participate meaningfully, but also to have built relationships with other experts and a reputation. The influence of the major OECD countries over the Inclusive Framework and the development of the Two Pillar Solution and its implementation may in time contribute to the weakening of the regimes. A full discussion of decision-making within the Inclusive Framework is beyond the scope of this article. It is simply observed here that arrangements perceived to be one-sided will create discontent, and once the groups of disaffected countries enhance their legal capacity, gridlock is a realistic outcome in a consensus-based process.

The expertise of the WTO and Appellate Body secretariats has been a technical strength of the WTO dispute settlement process, but ultimately became a political weakness.¹⁸⁴ The WTO dispute settlement panellists are formally independent but are commonly accused of being insiders with a pro-trade liberalisation outlook.¹⁸⁵ The OECD's role in drafting and building consensus around explanatory statements and official commentaries is likely to be more important, which are likely to carry considerable weight in international arbitration of Two Pillar Solution rules. What, if any, role the OECD Secretariat may have in the future assisting arbitration panels is unclear. The potential for the OECD Secretariat to influence tax arbitration may discourage some jurisdictions from accepting binding dispute settlement. The subject deserves further consideration by Inclusive Framework members.

This section has highlighted a number of issues with international tax dispute settlement under the Two Pillar Solution. Preliminary responses to only a couple of the issues have been suggested. All of the issues require further analysis. The aim has been to bring attention to the issues and hopefully stimulate further discussion. The difficulty of reaching agreement on reforms for ISDS suggests that the designers of dispute settlement systems for international taxation should endeavour to ‘get it right’ from the beginning.

VI. CONCLUSION

The prominent role that binding dispute settlement plays in the Two Pillar Solution is arresting when viewed alongside the limited role that MAP arbitration has played in the international tax system. The nature of Amount A necessitates some sort of multilateral process for its efficacy and administrative efficiency. However, questions need to be asked about a negotiation process that generated a dispute settlement system that has privileged corporate interests over public interests and ignored the political and policy lessons from other fields of international economic law. The interests of multinationals are advanced at the expense of State sovereignty by dispute settlement in relation to Amount A. Some commentators, however, celebrate these developments which they see leading to the fusion of horizontal disputes between tax authorities and vertical disputes between taxpayers and tax authorities within a State, which will in turn help to deal with multi-jurisdictional and multifaceted tax disputes more effectively.¹⁸⁶ The utility of the process for taxpayers does not ensure the legitimacy of the process. Tax arbitration impacts the broader public interest. Issues of transparency, arbitrator selection and capacity of smaller developing countries should not be neglected. In Pillar Two, where many States will have an interest in the GloBE rules being correctly applied, dispute settlement is likely to have a more limited scope, which will protect the large capital-exporting countries from legal challenges.

The role of mandatory binding arbitration in disputes of cross-border income taxation merits closer and more systemic attention. The enhancement of dispute settlement in the international income tax field is taking place at a time when States are reconsidering their consent to international adjudication of trade and investment disputes. The experience with dispute settlement in international trade and investment over the past two decades suggests that there are limits to what can and should be asked from the arbitration of international tax disputes. If tax arbitration decisions are decided against several jurisdictions, concerns about loss of sovereignty are likely to assume greater political salience. The influence exercised over the OECD/G20 BEPS Project and the Inclusive Framework by major developed countries suggests that the trade-offs inherent in the Two Pillar Solution rules and processes may be satisfactory for them today. Looking forward two decades, a changed economic environment may make today's rules less appealing, but the institutional structures may inhibit change in the direction(s) favoured by tomorrow's tax powers.¹⁸⁷ Many OECD documents embrace ‘common’ understandings about the tax treaties.¹⁸⁸ A similar clubbish-ness existed in the GATT, but that did not survive the appearance of a greater number of active and confident members within the WTO and extensive litigation.