The fragmentation of international investment and tax dispute settlement: A good idea?

Abstract

The international investment and tax law regimes are undergoing a process of significant reforms that seek to address existing shortcomings of the mechanisms used for the resolution of investment and tax treaty disputes. These reforms show that policymakers are gradually adopting a fragmented approach towards dispute settlement in both fields, with the establishment of different and unco-ordinated mechanisms. This article argues that, instead of fragmenting investment and tax dispute settlement, states should consider establishing a more unified and coherent framework in order to more adequately mitigate the concerns raised in each field.

1. Introduction

International investment law and international tax law are two of the fastest growing areas of international economic law, having led to the signing of thousands of mostly bilateral treaties that aim to facilitate cross-border trade and investment. The outcome of disputes under international investment agreements (IIAs or investment treaties) and double taxation treaties (DTTs or tax treaties) has severe implications for the development of international and national economic policies. In investment treaty disputes, arbitrators decide how public authorities may regulate in critical sectors and render awards against host states which in some cases can amount to a sizeable proportion of government budgets. For its part, the misinterpretation or disregard of tax treaty obligations creates uncertainty for taxpayers and in turn constitutes a significant barrier to cross-border transactions. Dispute resolution mechanisms in both fields, therefore, play a fundamental role to ensure the stability and predictability of the global economy.

The functioning of the current mechanisms to resolve disputes under DTTs and IIAs have come under increasing criticism in recent years. The predominant mechanism to resolve disputes under tax treaties, the Mutual Agreement Procedure (MAP), has been criticized for not always ensuring a satisfactory and timely resolution of the dispute and for failing to grant taxpayers participation rights. For the investment treaty regime, critics argue that investor-state dispute settlement (ISDS), inter alia, unduly restricts host states’ regulatory power, cannot guarantee arbitrators’ independence and impartiality and fails to ensure consistency in arbitral decisions.

Concerns have also been raised with respect to the interplay between DTTs and IIAs. Investment treaties contain substantive protection standards that can conflict with tax measures undertaken at the national level. Investment treaties do not, however, generally exclude taxation from their scope of application, meaning that they can cover tax measures aimed to raise revenue, eliminate double taxation or limit opportunities to engage in tax avoidance or evasion. Investors have brought tax-based ISDS claims in an increasing number of cases given the limits and shortcomings of the MAP. These claims can overlap with the subject matter covered by DTTs, creating uncertainties for tax and investment policymakers.

The critique of investment and tax dispute resolution mechanisms has triggered a reform process in both fields. Investment law reforms aim, among other purposes, to align investment protection with other policy objectives and to ensure greater consistency and coherence in the interpretation of investment treaty provisions. Tax law reforms focus on making the resolution of cases under the MAP more timely and effective, and on enhancing taxpayer participation in the procedure. Reform efforts to accomplish these objectives, however, reveal a fragmented approach towards dispute settlement, having led to the establishment of a plurality of unco-ordinated mechanisms to resolve investment and tax treaty disputes. In investment law, states are divided among maintaining the traditional ISDS model, introducing an investment court system, and replacing ISDS with state-to-state arbitration. In tax law, different arbitration mechanisms have been introduced as a supplement to the MAP in the Multilateral Convention to Implement Tax Treaty Related Measures to Prevent Base Erosion and Profit Shifting (the MLI) and the Council Directive (EU) 2017/1852 of 10 October 2017 (EU Dispute Resolution Directive). The MLI, for instance, introduces an independent-opinion arbitration and final-offer arbitration system for tax treaty disputes that cannot be resolved through the MAP.

This article examines whether the fragmentation of dispute settlement through investment and tax law reforms will address the weaknesses of each system. It demonstrates that, individually, the different dispute resolution mechanisms implemented in each regime do not adequately achieve the objectives behind the reforms. If anything, the proliferation of different dispute resolution systems will perpetuate instability and legal uncertainty not only for states but also for taxpayers/investors. Fragmenting dispute settlement can, in this respect, also trigger undesired practices, such as treaty and forum shopping. This article further shows that a lack of a clear definition on the relationship between IIAs and DTTs adds to the negative effects of fragmentation in investment and tax dispute settlement. As such, it argues that there is a need for a more unified and coherence dispute resolution framework in both fields and a more effective safeguard for avoiding overlaps between investment treaties and tax policymaking. The article concludes that, depending on its design, a multilateral regime could potentially fulfil these objectives.

2. Fragmenting dispute settlement mechanisms in international tax law

The international tax law regime is governed by more than 3,600 DTTs, most of which are based on the Organisation for Economic Cooperation and Development (OECD) Model Tax Convention. ¹ These treaties ‘serve several goals, including anti-double taxation of cross-border investment, prevention of excessive taxation, avoidance of tax evasion, cooperation in tax administration, and the exchange of information’. ² As the United Nations Conference for Trade and Development (UNCTAD) explains:

DTTs are not primarily focused on the unilateral tax rules in a given jurisdiction but rather on the interaction and overlap of these rules between two (or more) jurisdictions, each set of rules producing equitable and non-discriminatory results if taken in isolation. ³

Moreover, tax treaties generally apply only to direct taxation in the form of income, corporate profits and capital taxes.

Unlike investment treaties, DTTs do not provide taxpayers with direct access to international arbitration. Instead, disputes under these instruments are predominantly resolved through the (purely inter-governmental) MAP. This mechanism allows the competent authorities designated by the governments of the contracting states to resolve disputes arising from the actions of one or both contracting states’ tax administrations resulting in taxation not in accordance with the provisions of the treaty. A taxpayer who considers that they are being taxed inappropriately by one or both of the contracting parties may present the case to the competent authority of its resident state. ⁴

The MAP is thus of vital importance for taxpayers since it guarantees the proper application and interpretation of tax treaties. The MAP is, in this regard, critical to ensure the facilitation of cross-border trade by ensuring stability and certainty to international investors through the elimination of double taxation. However, the MAP has suffered from well-known criticism over the last few years. One drawback of this mechanism is that, as Chaisse notes, ‘the MAPs impose a relaxed responsibility on the competent authority, who just needs to “endeavor” to settle the controversy but is not “obliged” to settle the dispute’. ⁵ In cases where the competent tax authorities do not resolve the dispute, the taxpayer/investor will be subject to double taxation, which casts doubt on the effectiveness of DTTs.

Another criticism relates to the length of the MAP. As Lang and Owens observe, ‘a MAP is slow and the number of unresolved cases continues to grow, which has led to an increase in unrelieved double taxation’. ⁶ Several countries point to a significant accumulation of unresolved disputes, and the average duration of disputes has not shown any significant improvements. In this regard, the OECD’s statistics indicate that MAP cases closed in 2019 ‘lasted for 25 months (31 months for transfer pricing cases, 22 months for other cases)’. ⁷ This is particularly alarming if we consider that cases under the MAP have considerably increased over the past few years. The OECD’s statistics show that in 2019, around seven MAP cases were started every day, amounting to almost 2,700 new cases that year. ⁸

Another major concern with the MAP relates to its accessibility. Taxpayers suffering from double taxation or any other treatment inconsistent with DTTs are entitled to the MAP, but they have no rights of participation. ⁹ As Alschner explains, taxpayers ‘are not directly involved in the remainder of the proceedings, which take place behind closed doors’. ¹⁰ Thus, unlike investment treaty disputes, tax treaty disputes are still ‘politicized’. This leads to limited legal protection and a lack of transparency of proceedings. As will be explained in Section 4, the shortcomings of the MAP may be a reason for the increase in tax related claims before arbitral tribunals under IIAs.

International efforts to address the shortcomings of the MAP have been undertaken under the auspices of the OECD and by the EU. In 2013, the OECD launched the Base Erosion and Profit Shifting (BEPS) Project. The BEPS Project contains a ‘15 actions plan’ which addresses several concerns emanating from the vast network of tax treaties, such as treaty shopping and abuse leading to double non-taxation, corporate fiscal evasion, and harmful tax competition. ¹¹ The OECD was mindful that the implementation of the BEPS actions should not lead to unnecessary uncertainty for compliant taxpayers and unintended double taxation. It was therefore agreed that making dispute resolution mechanisms under tax treaties more effective and efficient should form an integral part of the BEPS Project. In this regard, the BEPS Project includes Action 14, which reads as follows:

BEPS Action 14 Make dispute resolution mechanisms more effective

Develop solutions to address obstacles that prevent countries from solving treaty-related disputes under MAP, including the absence of arbitration provisions in most treaties and the fact that access to MAP and arbitration may be denied in certain cases. ¹²

As part of Action 14, the OECD adopted in 2016 the Multilateral Convention to Implement Tax Treaty Related Measures to Prevent BEPS (also known as the MLI). ¹³ The MLI takes the form of an opt-in Convention, allowing parties to the DTTs to adopt the BEPS recommendations without having to renegotiate each relevant treaty. If both treaty partners consider their tax treaty as a ‘covered agreement’, the MLI modifies that treaty. ¹⁴ Thus, as Haslehner observes, ‘the MLI is best characterized as a framework agreement for the modification of bilateral tax treaties’. ¹⁵ As of July 2022, 99 jurisdictions have signed the MLI. ¹⁶

Among its objectives, the MLI aims at strengthening the MAP with provisions that stipulate specific timelines and ensure that the procedure is fully implemented in good faith and is more accessible to taxpayers. The MLI further introduces a supplementary (interstate) arbitration system for disputes that the MAP does not resolve. This mechanism is merely optional for signatories of the MLI. If the competent authorities of each contracting party have opted into the arbitration procedure, the decision is generally binding on them – unless both agree to another solution.

In parallel to the conclusion of the MLI, the EU adopted the EU Dispute Resolution Directive, which applies to the dispute resolution mechanisms provided in tax treaties signed between member states. The Directive similarly attempts to improve the MAP by introducing an (interstate) mandatory arbitration system. Moreover, unlike the MLI, the Directive allows member states to agree on a higher level of taxpayer participation in the proceedings. Instead of implementing a unified arbitration system, the MLI and the EU Dispute Resolution Directive offer distinct mechanisms that states can incorporate into their DTTs. These mechanisms are examined in turn.

2.1 Dispute settlement mechanisms under the MLI

It should first be noted that arbitration in international tax cases is not a new phenomenon. Several DTTs, in particular those conclude by the United States (US), already include arbitration as a method to resolve tax treaty disputes. ¹⁷ However, as with the MAP, a taxpayer/investor cannot directly access arbitration via tax treaties. Also, the issues that can be resolved through arbitration are limited. The arbitration clause serves, in other words, only as an extension of the MAP in cases where the dispute is not resolved through that procedure. ¹⁸ As the OECD has noted, the arbitration clause is an ‘additional dispute resolution technique which can help to … [ensure that] international tax disputes will to the greatest extent possible be resolved in a final, principled, fair and objective manner for both the countries and the taxpayers concerned’. ¹⁹ Moreover, if a court or administrative tribunal of either contracting party have already rendered a decision on a tax treaty dispute, the taxpayer/investor cannot resort to international arbitration. ²⁰

Part VI of the MLI introduces significant changes to international tax arbitration. ²¹ If a MAP is unsuccessful, the arbitration will become a mandatory second stage to resolve the dispute. The MLI also extends the scope of eligible matters that can be subject to arbitration. Further, an arbitral panel’s award (or decision) will be binding on the state party to the dispute. More innovatively, Article 23 of the MLI provides for two alternative arbitration methods. The first, ‘final offer’ arbitration (commonly known as ‘baseball arbitration’), is the default form of arbitration. The second is the ‘independent opinion’ arbitration, which states can select over ‘final offer’ arbitration.

Under the ‘final offer’ arbitration system, the competent authorities will present the arbitral panel with a proposed solution to the dispute and the arbitrators are only allowed to side with the position of one of the parties. In making the decision, the arbitrator will not provide reasons and cite legal authorities. ²² As such, the power of the arbitrators is severely restricted. This mechanism substantially differs from the ‘independent opinion’ arbitration method. This method is more akin to traditional arbitration proceedings, where the arbitrators will render a reasoned decision based on the parties’ arguments and evidence. This means that arbitrators will be called upon to interpret the relevant sources of law (treaty provisions and domestic legislation) and apply them to the facts of the case. In other words, arbitrators are not obliged to adhere to one of the proposed solutions presented by the competent authorities. Rather, they can decide independently on the solution to the case.

The role of the taxpayer in these two arbitration methods is limited. Under Article 19 of the MLI, the taxpayer/investor can refer the dispute to arbitration if the MAP case has not been resolved within a period of two years. However, as with the MAP, the arbitration proceedings will be under the control of the competent tax authorities.

Under Article 18 of the MLI, states can choose to incorporate one of these two arbitration mechanisms into their existing DTTs. On 28 July 2022, the OECD updated the ‘Arbitration Profiles’ of 31 jurisdictions that have decided to incorporate mandatory binding arbitration under Part VI of the MLI. ²³ These states are divided regarding their preference between the ‘final offer’ and ‘independent opinion’ arbitration mechanisms. For instance, the United Kingdom (UK), Luxembourg and Singapore have opted for the ‘final offer’ arbitration procedure, whereas Slovenia, Portugal and Papua New Guinea prefer the ‘independent opinion’ arbitration procedure.

2.2 Dispute settlement mechanisms under the EU Dispute Resolution Directive

The EU Dispute Resolution Directive also introduces substantial changes to international tax arbitration, but with respect to disputes under DTTs concluded between member states of the EU. ²⁴ If a MAP under an intra-EU DTT is unsuccessful, the Directive offers states the choice between different binding and non-binding dispute resolution methods. ²⁵ The Directive also extends the scope of eligible matters that can be subject to arbitration. However, the Directive goes beyond the MLI by allowing taxpayers/investors to take recourse to national courts to ‘unblock’ arbitration procedures. The Directive also enables the taxpayer to enforce the decision by resorting to national courts. Member states were expected to transpose the Directive’s set of rules into national law by the end of June 2019.

The Directive offers a variety of dispute resolution mechanisms that can be resorted to if the MAP fails. It adopts a ‘flexible’ approach towards ‘the choice of the method for dispute resolution’. Under Article 6, a taxpayer can, for instance, request that the competent authorities of the member states concerned establish an ‘Advisory Commission’ to resolve the dispute. The Advisory Commission should come to a decision by using ‘independent opinion’ arbitration. ²⁶ As explained above, this type of procedure requires that the arbitrators render a reasoned decision based on the parties’ arguments and evidence, examining the relevant sources of law (i.e., treaty provisions and domestic legislation) and apply them to the facts of the case.

Alternatively, the competent authorities can agree to set up an ‘Alternative Dispute Resolution Commission’. This Commission can also take the form of a committee that is of a permanent nature (a ‘Standing Committee’). Under Article 10, an Alternative Dispute Resolution Commission may apply, where appropriate, any dispute resolution processes or technique to solve the dispute in a binding manner, including the ‘independent opinion’ process or ‘the “final offer” arbitration process’. ²⁷

The Directive also encourages member states ‘to use non-binding alternative dispute resolution forums, such as mediation or conciliation, during the final stages of the mutual agreement procedure period’. ²⁸ By implication, therefore, the Alternative Dispute Resolution Commission can also resort to mediation or conciliation should the MAP be unsuccessful.

As to the role of the taxpayer under the Directive, unlike the MLI, the Directive provides the taxpayer with participation rights in addition to the right to initiate the dispute. It establishes a right for the taxpayer to provide the deciding panel with information, evidence or documents relevant to the decision. In addition, taxpayers may also appear (or be represented) before the panel.

2.3 Matching concerns with fragmentation?

From the above analysis, we can see that the tax law regime has adopted a fragmented approach towards dispute settlement. Indeed, the MLI and the EU Dispute Resolution Directive provide for a multiplicity of dispute resolution mechanisms and bodies that can be used as a supplement to the MAP, ranging from ‘final offer’ and ‘independent opinion’ arbitration procedures to mediation and conciliation through a permanent Alternative Dispute Resolution Commission. As such, it is likely that the landscape for resolving international tax disputes will significantly change in the near future.

For the time being, however, the brand-new procedures implemented through these instruments are in their infancy stage. This raises the question as to how these procedures will affect the resolution of tax treaty disputes and the position of taxpayers/investors. Differently put, it is unclear to what extent the fragmentation of dispute settlement will result in a system that will ensure an efficient and timely resolution of tax treaty controversies.

In the author’s view, the proposed fragmented dispute resolution system can create more problems than it solves. The mechanisms established under the MLI and the EU Dispute Resolution Directive leave, in this respect, a number of important issues unresolved. Take, for example, the ‘final offer’ arbitration procedure. As one author writes, this dispute resolution method arguably provides ‘a tantalizing alternative to conventional arbitration … because it appears to be more prone to productive negotiations, faster, cheaper [and] helps to safeguard long-term relationships’ between states. ²⁹ This process can, however, bring a number of negative externalities. One relates to the limited authority given to arbitrators. There is ultimately no ‘decision’ to be rendered in this procedure since the arbitrators need not interpret and apply treaty provisions based on an independent assessment of the parties’ views. Accordingly, as Petruzzi notes, ‘final offer’ arbitration cannot create a precedent, ³⁰ implying that it does not provide guidance for future arbitrators deciding disputes under the same DTT.

The ‘final offer’ arbitration mechanism could also lead to challenges under national arbitration laws. Limiting ‘the parties’ rights to a reasoned award’, Gupta observes, ‘is prejudicial to an award’s validity’. ³¹ The parties’ right to have a reasoned award forms part of the mandatory provisions of most modern arbitration seats around the world. Article 1482 of the French Civil Code of Procedure, for instance, provides that ‘the arbitral award shall succinctly set forth the respective claims and arguments of the parties and … shall state the reasons upon which it is based’. ³² Similarly, Section 611(2) of the Austrian Code of Civil Procedure provides that a failure to provide decisive reasoning can be a ground for setting aside an arbitral award. In applying this provision, the Austrian Supreme Court recently set aside an arbitral award partially due to the violation of the procedural ordre public on the grounds that the arbitral tribunal did not sufficiently reason its award. ³³

The ‘independent opinion’ arbitration mechanism could address these concerns since it allows the arbitrators to rule on the entirety of the dispute, considering the facts and arguments as presented by the parties. However, as Mooij aptly notes, the independent opinion as implemented in the MLI and EU Dispute Resolution Directive is ‘silent on who shall administer arbitrations, and so are their official explanations nor, for that matter, are there more than a minimal number of existing bilateral tax treaties that address the issue of administration’. ³⁴ This will bring new challenges regarding the support and conduct of the arbitration, such as constituting the arbitral tribunal, ensuring arbitrators independence and impartiality, and determining the applicable procedural rules (e.g., submission of evidence and organization of hearings).

Developing fair and effective rules to ensure independence and impartiality is fundamental to the legitimacy of the tax treaty regime and each member of the arbitral tribunal. Concerns over how to ensure independence and impartiality are also reflected in the MLI, which provides that ‘[e]ach member appointed to the arbitration panel must be impartial and independent’ at the time of accepting the appointment and during the proceedings. ³⁵ The MLI further clarifies that arbitrators must be free of any relationship with ‘the competent authorities, tax administrations, and ministries of finance of the Contracting Jurisdictions and of all persons directly affected by the case (as well as their advisors)’. ³⁶ The MLI, however, does not contain a provision with disclosure obligations and the possibility for the parties to challenge an arbitrator.

A further unresolved issue with the ‘independent opinion’ procedure relates to the fiscal sovereignty of states. Hearson and Tucker claim that ‘[u]nder mandatory binding tax arbitration, states cede sovereignty over the interpretation of international tax agreements to panels of transnational tax adjudicators of states’. ³⁷ Sidhu similarly observes that ‘states are extremely protective of their fiscal sovereignty and unwilling to subject their taxing powers to adjudication’. ³⁸ Cruz believes that if arbitral decisions conflict with domestic sovereignty regulations, competent authorities may have no incentive to conclude MAP cases or to avoid treaty interpretations that undermine the other state and the interests of taxpayers. ³⁹

The ‘independent opinion’ procedure can also give rise to interpretative inconsistency concerns. Given the existence of thousands of DTTs containing similar language, different arbitral panels can be asked to apply and interpret the same investment treaty or different treaties with similar provisions. However, in their current form, the arbitration mechanisms of the MLI and the Directive do not provide a system of precedent that constrains arbitrators to follow the interpretations adopted by previous panels. This can lead to inconsistent interpretations. Wijnen explains this concern as follows:

even if the arbitration were one day to become a regular judicial phenomenon, there is little chance that arbitration on a per treaty basis would ever result in a uniform interpretation of tax treaties. As arbitration commissions are not bound by the laws of the two treaty partner states, their decisions are not automatically taken as a precedent by the courts of these states. ⁴⁰

The lack of uniformity in treaty interpretation will undermine the legitimacy of the tax law regime and will make it difficult for states and taxpayers/investors to ascertain the exact scope of treaty commitments.

Moreover, the arbitration mechanisms implemented through the MLI and the EU Dispute Resolution Directive will remain under the exclusive control of the competent tax authorities. The Directive gives member states room for conferring participation rights upon taxpayers, but the exact scope of those rights is unclear. This is at odds with the fact that the taxpayer should initiate the arbitration procedure since he is the ultimate beneficiary of the proceedings suffering from double taxation. As Perrou rightly remarks, ‘at the current level of development of international (economic) law and human rights law, [the absent taxpayer] can no longer be justified’. ⁴¹

The fragmentation of dispute settlement can also instigate treaty shopping practices. Rosenbloom described this phenomenon as:

the practice of some investors of “borrowing” a tax treaty by forming an entity (usually a corporation) in a country having a favourable tax treaty with the country of source -that is, the country where the investment is to be made, and the income in question is to be earned. ⁴²

Treaty shopping is mainly associated with the substantive benefits of tax treaties. With the forthcoming diversification of dispute settlement mechanisms in DTTs, treaty shopping can become a tool that taxpayers can also employ to ‘shop’ between more favourable procedural protections that would have otherwise been unavailable.

For example, a corporation (LUCo) resident in Luxembourg (the home country) may own a corporation (CACo) in Canada (the source country). Suppose that the Luxembourg-Canada DTT does not provide for arbitration, but the DTT between Luxembourg and the US allows taxpayers to resort to ‘independent opinion’ arbitration. LUCo could qualify for the procedural benefits under the Luxembourg-US DTT by forming a corporation (USCo) in the US (the third country) or by transferring the stock of CACo to USCo.

By supplementing the MAP with a mandatory arbitration mechanism, the OECD and the EU certainly strive towards establishing a more efficient and effective resolution of tax treaty disputes. Yet, the creation of different and potentially overlapping arbitration mechanisms may not ultimately be the way forward to achieve that purpose. Several deficiencies still remain in each proposed mechanism, resulting in a higher level of uncertainty for both taxpayers and states. The aim should be to achieve more predictability, and it is doubtful whether fragmentation is the right approach to that end. A similar problem is palpable in the investment law regime.

3. Fragmenting dispute settlement mechanisms in international investment law

The international investment regime is governed by over 3,000 IIAs, including bilateral investment treaties (BITs), free trade agreements (FTAs) and multilateral investment agreements. ⁴³ These treaties aim to promote foreign direct investment. ⁴⁴ They provide investors with an unprecedented level of substantive and procedural protections but offer no reciprocal rights for states wishing to preserve regulatory space. ⁴⁵ Substantively, investment treaties generally require host states to pay compensation in the event of an expropriation, provide fair and equitable treatment (FET) as well as complete protection and security, provide equal treatment for foreign and domestic investors (national treatment) and treat all foreign investors alike (most-favoured-nation treatment (MFN)). ⁴⁶

Procedurally, unlike DTTs, investment treaties allow investors to directly challenge policy measures that may affect their investments before arbitral tribunals and claim for high compensation amounts. ⁴⁷ Investment treaties contain ISDS provisions that offer investors the possibility to resort to different arbitral institutions for the administration of the process, including the International Centre for the Settlement of Investment Disputes (ICSID), ad hoc tribunals established under the United Nations Commission on International Trade Law (UNCITRAL) Arbitration Rules, and the International Chamber of Commerce (ICC). ⁴⁸

The preferred mechanism for the resolution of investment disputes is ICSID. ⁴⁹ The ICSID regime, unlike the other arbitral fora where the award can be set aside before national courts at the seat of the arbitration, is largely self-contained. The review of the award is, thus, limited to the ICSID annulment system and the narrow grounds outlined in Article 52(1) of the ICSID Convention. This Convention contains stricter grounds for annulment of the award than those provided in the New York Convention on the Recognition and Enforcement of Foreign Arbitral Awards, which commonly governs non-ICSID arbitration proceedings. ⁵⁰

The number of ISDS claims reached 1,190 at the end of 2021. ⁵¹ As discussed in Section 4, these claims include arbitrations arising from taxation measures adopted by host states, such as windfall tax, value-added tax, income tax, and import taxes. The unprecedented rise in ISDS claims against states has put into question the legitimacy of this system and the treaties from which its jurisdiction and substantive rules largely derive. ⁵² Critics argue that the regime unduly restricts host states’ regulatory policy space, cannot guarantee arbitrators’ independence and impartiality, fails to ensure consistency between decisions, lacks transparency, and leads to overly long and expensive proceedings. ⁵³

Criticism surrounding the independence and impartiality of arbitrators is ‘directed against the central feature of party appointment, and the propriety of connections between arbitrators and parties, the issues of multiple appointments, double-hatting, and issue conflict, and implicit pro-investor [or pro-state] bias’. ⁵⁴ Party appointment of arbitrators can at times be perceived as a ‘moral hazard’ and become problematic. ⁵⁵ The reason being that an arbitrator appointed by the state, for instance, may have an incentive to decide in its favour with the objective to express loyalty and obtain reappointments. Repeat appointments in turn risk increasing an arbitrator’s tendency to decide in favour of the party making such appointments. Double-hatting is another growing concern. Some of the arbitrators who decide ISDS claims also periodically serve as counsel in other investment arbitration cases. The fear here is that an arbitrator’s decision might be influenced by arguments they wish to make in a case where they are litigants. These situations may create the appearance of conflict. ⁵⁶

Another notable concern is interpretative inconsistency, considered as ‘one of the most salient problems for governments as well as commentators’. ⁵⁷ It goes to the very heart of the backlash against the regime. As in the tax system, there is no system of precedent in investment arbitration that constrains arbitrators to follow prior awards. As such, arbitrators tend to adopt different interpretations of provisions contained in the same investment treaty or of identical or similarly worded provisions contained in different investment treaties. A recent example of inconsistency in treaty interpretation includes decisions where tribunals have taken opposite views as to whether dual nationals should be entitled to treaty protection. ⁵⁸

In response to these criticisms, an increasing number of states are reviewing and amending their IIAs. ⁵⁹ This reform process is frequently framed in terms of the need to mitigate the asymmetry prevalent in investment treaties by aligning investment protection with other state interests. In this regard, states are incorporating provisions in investment treaties to safeguard the right to regulate, impose obligations upon investors and limit the scope of investment protection standards. ⁶⁰ Investment law reforms also cover dispute resolution mechanisms. Although most investment treaties maintain the traditional ad hoc ISDS system, an increasing number of states are resorting to other alternatives, thereby adopting a fragmented approach towards dispute settlement. Given space constraints, we will only focus on the most salient reforms that have been implemented, namely the establishment of an investment court system, the replacement of ISDS with interstate arbitration mechanisms and the introduction of exhaustion of local remedies requirements.

Some of these reforms are also being discussed under the auspices of the UNCITRAL Working Group III. This Group was given a broad mandate to: (i) identify and consider concerns regarding ISDS, (ii) consider whether reform is desirable and, if so, (iii) develop any relevant solutions to be recommended to the UNCITRAL Commission. ⁶¹ The Group has also identified concerns pertaining to consistency, coherence, predictability and correctness of decisions by arbitrators, concerns pertaining to their independence and impartiality, and concerns pertaining to cost and duration of ISDS cases. In April 2019, the Working Group agreed that reform was necessary and began to discuss reform options, ⁶² which are currently being developed through a draft work plan. ⁶³

3.1 The investment court system

In Europe, mobilization against ISDS has been exceptionally high, to the point that the EU Trade Commissioner dubbed ISDS ‘the most toxic acronym in Europe’. ⁶⁴ In March 2018, the Court of Justice of the European Union (CJEU) held in the Achmea case that the ISDS clause in the Netherlands-Slovakia BIT is incompatible with EU law. ⁶⁵ Following up on the legal consequences of this ruling, in January 2019, member states issued declarations in which they agreed to terminate their intra-EU BITs. ⁶⁶ As a result, potential alternatives for the resolution of intra-EU investment disputes are under discussion.

At the same time, the EU has developed an investment court system (ICS) to hear claims under treaties concluded with non-EU states. ⁶⁷ The ICS was included in the Comprehensive Economic and Trade Agreement (CETA), concluded between the EU and its member states and Canada. ⁶⁸ This system, which was recently ‘Europeanized’ by the CJEU, ⁶⁹ replaces the traditional ISDS model found in most investment treaties. The ICS has also been included in other trade agreements concluded with other states, including Singapore, Vietnam and Mexico. ⁷⁰

The ICS represents a significant departure from the long-standing ad hoc ISDS model of party-appointed arbitrators. It provides for a standing mechanism for the settlement of disputes with two levels of adjudication: a first instance and an appellate tribunal. If a dispute has not been resolved through consultations, an investor may then proceed to arbitration before the First Instance Tribunal. The Tribunal’s remedies are confined to either monetary damage and any applicable interest, or restitution of the property. An award by the first instance tribunal may be appealed to the Appeal Tribunal within ninety days. The Appeal Tribunal will review the awards based on (i) errors in the application or interpretation of the law, (ii) manifest errors in the appreciation of facts, including the appreciation of relevant domestic law, and (iii) the grounds set out in Article 52(1)(a)–(e) of the ICSID Convention. ⁷¹

With respect to the appointment of arbitrators, investors have no say in the determination of the ICS’s members deciding their claim. This is so with respect to the selection process of the members of the First Instance Tribunal and the Appellate Tribunal. Members of the Appellate Tribunal will be selected by a Joint Committee of the treaty parties after the entry into force of CETA based on a permanent roster of fifteen judges. The roster will be made up of five EU nationals, five Canadian nationals, and five third party nationals. ⁷² They are appointed for two five-year terms and a four-year term. CETA also prevents arbitrators from acting as counsel or as party-appointed expert or witness in any pending or new investment dispute under CETA or any other international agreement. ⁷³

In its thirty-eighth session, the UNCITRAL Working Group IIII also proposed the inclusion of an appellate mechanism in investment treaties as a possible reform of the traditional ISDS system. This suggestion was contained in various proposals submitted by governments in preparation for the deliberations on the third phase of the mandate of the Group. As part of the discussions, the Group examined grounds of appeal and standard of review as a central feature to establish the appellate mechanism. ⁷⁴

3.2 State-to-state arbitration

Another unique solution implemented by states to address the concerns of the current system has been to replace ISDS with state-to-state dispute settlement (SSDS). Several states have already opted for this reform option. UNCTAD’s IIA Mapping Project indicates that currently 296 out of 2,584 mapped IIAs only provide for SSDD. ⁷⁵ The most recent examples include the EU-UK Trade and Cooperation Agreement, Brazil’s Cooperation and Facilitation Investment Agreements, and the US-Mexico-Canada Agreement (USMCA) with respect to the US and Canada.

For instance, under Chapter 31 of the USMCA, a dispute settlement process is available to deal with disputes among USMCA parties ‘regarding the interpretation and application of the agreement or whenever a party state believes that an actual or proposed measure of another USMCA state is or would be inconsistent with that state’s obligations under the USMCA’. ⁷⁶ The SSDS in the USMCA provides that if a dispute regarding ‘the interpretation and application of’ the agreement arises ‘the complaining Party may select the forum to settle the dispute’. ⁷⁷

Similarly, the Brazil-Malawi BIT allows the treaty parties to choose ‘arbitration mechanisms’ with the agreement of a Joint Committee that should first attempt to settle the dispute:

Before initiating an arbitration procedure, any dispute between the Parties shall be assessed through consultations and negotiations between the Parties and previously examined by the Joint Committee … If the dispute cannot be resolved, the Parties to the exclusion of the investors may resort to arbitration mechanisms between States, which are to be agreed upon by the Joint Committee, whenever the Parties find it appropriate. ⁷⁸

Other investment treaties provide for ad hoc arbitration between the contracting parties or adjudication before the International Court of Justice. The Germany-Pakistan BIT, a treaty often described as the first BIT to be concluded, provides a typical formulation:

In the event of disputes as to the interpretation or application of the present Treaty, the Parties shall enter into consultation for the purpose of finding a solution in a spirit of friendship. (2) If no such solution is forthcoming, the dispute shall be submitted (a) to the International Court of Justice if both Parties so agree or (b) if they do not so agree to an arbitration tribunal upon the request of either Party. ⁷⁹

Some investment treaties submit the disputes to regional courts. For example, the Investment Agreement for the COMESA Common Investment Area states:

Any dispute between Member States as to the interpretation or application of this Agreement not satisfactorily settled through negotiation within 6 months, may be referred for decision to either: (i) an arbitral tribunal constituted under the COMESA Court of Justice in accordance with Article 28(b) of the COMESA Treaty; or (ii)an independent arbitral tribunal; or (iii) the COMESA Court of Justice sitting as a court. ⁸⁰

SSDS is not a new phenomenon. Most investment treaties typically contain SSDS settlement provisions in addition to, rather than in the place of, ISDS clauses. ⁸¹ SSDS clauses offer states a parallel and autonomous procedural right to espouse a treaty claim on behalf of its injured national. As Kulick explains, SSDS provisions can lead to the following category of state-to-state arbitration: ‘(2) diplomatic protection claims, in which the home state introduces state-to-state dispute settlement on behalf of the investor for violation of the IIA’. ⁸² It should be noted that investment treaties do not generally prioritize ISDS provisions over SSDS provisions. ⁸³ This means that the home state can initiate an arbitration even in the presence of an ISDS provision in the treaty.

3.3 Local remedies

Some BITs also include the requirement of the exhaustion of local remedies. UNCTAD’s IIA Mapping Project indicates that currently 89 out of 2,584 mapped IIAs require foreign investors to first try to resolve the dispute in national courts before submitting the dispute to arbitration. ⁸⁴ A requirement to exhaust local remedies recognizes that domestic courts are the most appropriate jurisdiction to hear a grievance arising in the public domain. ⁸⁵ The Albania-Lithuania BIT, for instance, provides that:

In the event of a dispute the Contracting Party in whose territory the investment was made shall be notified in writing, including detailed information by the investor … If such a dispute cannot be settled amicably within six months from the date of the written notification provided in paragraph 1, and domestic judicial and administrative remedies have been exhausted, the Contracting Party or the investor shall be entitled to submit the dispute [to arbitration]. ⁸⁶

As another example, the 2018 India-Belarus BIT contains a timeframe for the presentation and duration of the claim. The treaty states that:

a disputing investor must first submit its claim before the relevant domestic courts or administrative bodies of the Defending Party … within two (2)year(s) from the date on which the investor first acquired, or should have first acquired, knowledge of the measure’ damaging the investment. ⁸⁷

The treaty then requires the investor to pursue domestic remedies related to the measure ‘for at least a period of five years’. ⁸⁸

The ICS, SSDS and domestic remedies represent only a taste of the different mechanisms that states are including in their treaties in view of the concerns pertaining to the functioning of the traditional ISDS system. States are also adopting alternative dispute resolution (ADR) methods, such as conciliation and mediation. As has been noted ‘[t]he recent reforms of treaties signed by states, either in the form of an investment chapter of an FTA or as stand-alone BITs, show that mediation/conciliation is slowly gaining attention and traction in treaty language’. ⁸⁹ UNCTAD’s IIA Mapping Project indicates that currently 627 out of 2,584 mapped IIAs contain either voluntary or compulsory mediation/conciliation mechanisms. The UNCITRAL Working Group III is also proposing mediation as a method of resolving ISDS disputes. ⁹⁰

3.4 Matching concerns with fragmentation?

The foregoing analysis shows that, similar to the tax law regime, the investment law system adopts a fragmented approach towards dispute settlement. The question now also arises as to whether the establishment of different and unco-ordinated dispute settlement mechanisms is, in fact, the way forward to address the different concerns related to the traditional ISDS system.

The creation of an ICS with an appellate mechanism has the potential to mitigate some of the criticisms, such as the independence and impartiality of arbitrators. The establishment of a permanent roster of judges appointed by the CETA Joint Committee can alleviate concerns stemming from the role of party autonomy in the appointment of arbitrators. This is particularly so with respect to reappointments as well as conflicts of interests and the perception of bias that arise from this practice. The CETA’s rule that arbitrators cannot work as counsel or expert in another proceeding is also a positive development to address the problem of double-hatting. Moreover, ‘[d]epending on the grounds of appeal’, Giorgetti writes, the appellate mechanism could ‘scrutinize how double-hatting, issue conflicts or contacts between arbitrators and parties are dealt with in first-instance arbitrations and set aside or annul arbitral awards if inappropriate behaviour is detected’. ⁹¹ However, it can be argued that the ICS would invert problems about the independence and impartiality of arbitrators. This is because a nomination system controlled by states can lead to a court populated by ‘pro-State’ judges. ⁹² This problem could be addressed by replacing the system of state party appointment with one where judges are appointed by an institution or an international organization.

The ICS can also deal with concerns relating to inconsistent interpretations. As previously explained, international investment law is composed of a decentralized and unco-ordinated network of thousands of, mostly bilateral, investment treaties. Many countries in the world have negotiated these instruments on the basis of treaty models, and there is a high degree of similarity in the wordings of treaty provisions. As such, different arbitral panels have been asked to apply the same investment treaty or different treaties with similar provisions. With a standing two-tier system, a MIC would ensure predictability and consistency in the interpretation of treaty provisions, thereby moving towards a precedent-based system. ⁹³ It would, in other words, create an opportunity to establish authoritative guidance on the application of the treaty. However, as Zarra notes:

it is likely that the goal of coherence cannot be reached through the establishment of an investment court which operates only within the framework of a single treaty: this might, at most, ensure consistency in the application of such a treaty, but surely the court will be not able to ensure consistency with regard to other treaties, considering that—in lack of a principle of stare decisis in international investment law—different tribunals are potentially free to interpret other treaties’ provisions as they want. ⁹⁴

Moreover, in its current form, the ICS does not address concerns about the asymmetry prevalent in the system. As mentioned above, most investment treaties provide foreign investors with an unprecedented level of substantive and procedural protections but offer very little safeguards to host countries. In this respect, arbitrators do not generally consider arguments by states regarding their policy powers to regulate, nor do they value other non-investment interests of states when deciding investment disputes, such as human rights, the environment and labour standards. ⁹⁵ The ICS still prioritizes the rights of foreign investors over the public interests of states. It offers a one-sided dispute resolution mechanism, meaning that states are not allowed to bring claims against investors for their misconduct. A SSDS mechanism could mitigate these concerns.

SSDS can serve to align investment protection with other public interests, showing more respect for state sovereignty. ⁹⁶ In this connection, states can apply restraints with respect to the claims they are willing to pursue on behalf of investors and the type of government measures that can be challenged. States can also exert more control over the resolution of the dispute. This has the potential to prevent investment disputes as it creates a stronger incentive for investors to negotiate in collaboration with their home state. However, as an author has explained:

State-state arbitration would not resolve concerns about lack of consistency, coherence, predictability and correctness of arbitral decisions, unless major changes were also made to arbitral procedures. Where there was standing consent to state–state arbitration, power asymmetries could still apply, and the chilling effect of diplomatic pressure on governments to change their proposed actions might even intensify. ⁹⁷

In addition, in terms of procedural efficiency, one may question the need for diplomatic protection claims, especially if a related ISDS claim is pending. The international community provided good reasons to create a system that allows investors to bring direct claims against host states, a system that usually offers greater resources to take up claims.

The fragmentation of dispute settlement mechanisms can also trigger the practice of treaty shopping. This phenomenon is also present in international investment law. Access to virtually all investment treaties depends on the claimant investor qualifying as a national of one of the contracting parties. However, many investment treaties along with the ICSID Convention, are premised on a rather rudimentary set of provisions defining who is an eligible national or investor. With the support of investment jurisprudence, broad definitions of investor have enabled corporations and individuals to create a diversity of nationality with the purpose of gaining access to ISDS or benefiting from the most liberal treaties. Investors can, therefore, avoid resorting to undesirable forums, such as the ICS or SSDS, by structuring their investments through intermediate companies based in home state jurisdictions that have signed investment treaties that maintain the ad hoc ISDS mechanism. ⁹⁸

In short, similar to the conclusion reached with respect to the tax law regime, the inclusion of different and unco-ordinated dispute resolution mechanisms in investment treaties may not be the way forward to resolve the existing deficiencies of ISDS. An essential objective of a successful dispute resolution system should be to achieve some level of predictability and certainty and to consistently deliver comprehensive justice. It is doubtful whether the different mechanisms implemented in each regime will achieve that objective, at least individually. As will be shown, the potential overlap between disputes subject to investment and tax treaties further illustrates the potential side effects of fragmentation.

4. The interplay and overlap between investment and tax treaties

As O’Brien and Brooks explicate, ‘[t]ax treaties and IIAs have much in common’. ⁹⁹ These instruments ‘share the same purpose of facilitating FDI, … provide similar legal protections, such as prohibition of discriminatory treatment of non-nationals [and] are intended to create security and predictability’ for investors. ¹⁰⁰ IIAs and DTTs have also ‘proliferated in tandem during the recent period of intensified globalization’. ¹⁰¹ IIAs offer, however, a larger scope of protection for investment than DTTs. ¹⁰² As discussed above, tax treaties do not provide investors/taxpayers with direct access to dispute resolution and only deal with the allocation of taxing rights between contracting parties over certain types of income and capital gains. Investment treaties offer expansive substantive protections in respect of investments that generate that income and enable investors/taxpayers to bring direct claims against host states.

Moreover, most IIAs do not exclude taxation from their scope of application and are silent on their relationship with DTTs. ¹⁰³ This means that investors can be protected from tax-related measures adopted by host states that violate the IIA’s substantive protections, including measures that ‘may simultaneously fall within the scope of a DTT as well as an IIA between the relevant countries’. ¹⁰⁴ More recent investment treaties contain tax carve-out provisions, ¹⁰⁵ which exclude tax measures from all or certain investment protection standards and attempt to prevent inconsistencies relating to a taxation measure between IIAs and DTTs. However, these treaties do not generally define what is meant by ‘taxation measure’, nor do they explain who (investment tribunal or domestic tax authorities) should solve potential inconsistencies. ¹⁰⁶

The better protection offered by IIAs has resulted in a multiplication of tax disputes before investment treaty tribunals. According to UNCTAD, between 1987 and 2021, ‘investors have challenged tax-related measures in 165 ISDS cases based on IIAs’. ¹⁰⁷ These cases involve different measures, including regulatory changes to feed-in tariffs for renewable energy production, withdrawal of VAT subsidies, increase in windfall profit taxes and royalties, the initiation of tax investigations or audits, and the imposition of capital gain taxes. ¹⁰⁸

ISDS claims involving domestic tax policies have the potential to overlap with the subject matter covered by DTTs and MAPs. An illustrative recent example is the Cairn v. India case. ¹⁰⁹ This case arose out of India’s decision to retrospectively amend its income tax laws and impose a tax liability of US$1.6 billion on Cairn India Ltd for its failure to deduct withholding tax on capital gains resulting from a series of restructuring transactions that took place among the Cairn group in 2006. ¹¹⁰ Cairn UK initiated UNCITRAL arbitration proceedings under the UK-India BIT, claiming that India’s measures leading to the imposition of the retroactive tax breached, among others, its obligation to accord Cairn UK and its investment fair and equitable treatment.

India made several jurisdictional objections, including that challenges to its ‘tax legislation and policy are excluded from the scope of the BIT and are not arbitrable’:

tax disputes are not capable of being resolved by arbitration under the BIT in light of an implied exception to the scope of application of the BIT, and of the fact that the Respondent and the United Kingdom have in fact specifically agreed that tax disputes should be settled in accordance with the procedure prescribed in the contemporaneous [double taxation avoidance agreements]. ¹¹¹

India relied, in this respect, on the UK-India DTT, ‘which does not provide for arbitration, but rather for a mutual agreement procedure involving consultations between the taxation authorities of the two States’. ¹¹² According to India:

the advancement of [tax] claims under the BIT is incompatible with the [DTT], in which the Respondent and the UK seek to ensure the “avoidance of double taxation and the prevention of fiscal evasion with respect to taxes on income and capital gains”. ¹¹³

In other words, since the measures adopted by India are regulated by the DTT, ‘the BIT should be read so as to exclude such matters from its scope’. ¹¹⁴ The absence of a tax carve-out in the BIT defining the relationship between this treaty and the DTT did not alter this conclusion, India added, ‘because at issue here is the existence of general limits to the scope of protection of investment treaties which exist even if they are not made explicit’. ¹¹⁵ Cairn UK should have, therefore, resorted to the MAP provided in the DTT, instead of challenging India’s tax measures under the BIT.

The Tribunal disagreed with India. It observed that the UK-India DTT and the UK-India BIT ‘govern different subject-matters’ and that the BIT ‘does not expressly specify that [it] should be considered to be incompatible with’ the DTT. ¹¹⁶ The Tribunal noted, in this respect, that, unlike the ISDS provision in the BIT, the MAP ‘does not purport to provide a dispute resolution mechanism for situations in which an investor of one of the Contracting States considers that the host State has violated his rights as an investor’. ¹¹⁷ The Tribunal further found that the BIT did not contain a provision preventing the investor from submitting arbitration claims relating to tax measures that can potentially fall within the scope of the DTT. The Tribunal upheld jurisdiction over the dispute and held that India had failed to respect its obligations under the BIT, in particular, the FET standard. ¹¹⁸ India was ordered to pay Cairn UK over US$1.2 billion in compensation. ¹¹⁹

The ConocoPhillips v. Vietnam case is another example of how investment treaty protection can conflict with rights and obligations under DTTs. This case also relates to capital gains tax on restructuring of assets. In 2012, ConocoPhillips UK (a UK subsidiary of the US energy giant ConocoPhillips) sold two of its entities (ConocoPhillips Gama Limited and ConocoPhillips Cuu Long) to UK-based Perenco Overseas Holding. The only assets held by ConocoPhillips Gama and Cuu Long were ConocoPhillips’s oil interests in Vietnam. It was reported that ConocoPhillips sold the companies for US$1.29 billion, making a profit of US$896 million. ¹²⁰

Under the terms of the UK-Vietnam DTT, ‘capital gains generated from transactions involving shares deriving their value from immovable property situated in one of the contracting states may be taxed in the jurisdiction where the property is located’. ¹²¹ The Vietnamese tax administration interpreted the DTT as granting the state the right to tax capital gains on the transaction since it derived its value exclusively from oil interests located in Vietnam. ¹²² Based on the current tax rate in Vietnam, ConocoPhillips would have to pay an estimated US$179 million to the Vietnamese government for its capital gain. ConocoPhillips refused to pay this tax, arguing that the sale was between two UK entities with no taxable presence in Vietnam.

In 2015, Vietnam signalled its intention to tax the transaction. In a move designed to prevent the Vietnamese government from collecting the capital gains tax, ConocoPhillips and Perenco initiated UNCITRAL arbitration proceedings against Vietnam under the UK-Vietnam BIT. ¹²³ On 20 January 2020, the journal Finance Uncovered reported that ConocoPhillips has settled the case with the government, noting that ‘the US oil giant, has finally paid tax to Vietnam on a $896m gain from the sale of two oil fields in 2012 – marking a significant climbdown amid embarrassing legal action and international critic’. ¹²⁴ Although the settlement of the dispute has been confirmed, the exact terms of the settlement remain undisclosed.

Had the dispute proceeded, Vietnam would have likely raised a jurisdictional objection on grounds like those invoked in Cairns v. India, arguing that any disagreement between Vietnam and ConocoPhillips as to the payment of the capital gains tax should have been resolved through the UK-Vietnam DTT. It is also probable that tribunal would have rejected the objection on the basis that the UK-Vietnam BIT does not require that investors resorts to the DTT to challenge tax related measures.

A final case worth mentioning is Schooner v. Poland. ¹²⁵ This case involved a BIT claim arising out of an investment made by two US companies in the mid-1990s in a newly privatized Polish state enterprise, Kama Foods, an oil and margarine manufacturer. Kama Foods undertook to receive the payment of management services fees, training and know-how to the company. For fiscal years 1994 to 1997, Kama Foods recorded the payment of these management fees as tax-deductible for tax assessment purposes. As a result of a series of inspections conducted in 1997, the Polish tax authorities took a series of tax enforcement measures that disallowed certain deductions that had been taken by Kama Foods, leading the company to become insolvent.

On 31 March 2011, the investors instituted arbitration proceedings under the ICSID Additional Facility Rules pursuant to the US-Poland BIT. The investors argued that, through its tax measures, Poland had violated the expropriation, FET, and full protection and protection standards (FPS) of the Poland-US BIT as well as its provisions relating to the free transfers of investments. Poland raised several jurisdictional objections. In particular, Poland argued that ‘the entire Tax Claim is covered by the tax exception provided in Article VI of the Treaty read in conjunction with Article 22 of the Poland – United States Double Tax Treaty (“DTT”) and is, therefore, outside the jurisdiction of this Tribunal’. ¹²⁶ Article VI of the BIT is a tax carve-out provision which reads as follows:

1. With respect to its tax policies, each Party should strive to accord fairness and equity in the treatment of, and commercial activity conducted by, nationals and companies of the other Party.

2. Nevertheless the provisions of this Treaty, and in particular Article IX and X, shall apply to matters of taxation only with respect to the following:

(a) expropriation, pursuant to Article VII;

(b) transfers, pursuant to Article V; or

(c) the observance and enforcement of terms of an investment agreement or authorization as referred to in Article IX(1)(a) or (b),

to the extent that they are not subject to the dispute settlement provisions of a convention for the avoidance of double taxation between the two Parties, or have been raised under such settlement provisions and are not resolved within reasonable period of time. ¹²⁷

Poland first argued that ‘the phrase “matters of taxation” in Article VI(2) should be defined broadly as referring to all issues related to the process or system of imposing and charging taxes’. ¹²⁸ As such, according to Poland, the Tribunal did not have jurisdiction over the claimants’ FET and FPS claims since the tax measures adopted by the state fell within the ambit of that provision. The Tribunal majority agreed, holding that ‘matters of taxation’ include the ‘assessment and collection of taxes’, which is the type of measure that triggered the claimants’ BIT claims. ¹²⁹

Second, Poland argued that the Tribunal did not have jurisdiction over the claimants’ remaining claims, expropriation and denial of free transfer, ‘because the Claimants did not resort to the “dispute settlement provisions of a convention for the avoidance of double taxation” before initiating this arbitration as required under Article VI(2) of the BIT’. ¹³⁰ Article 22 of the Poland-US DTT provides for the MAP. According to the Respondent, ‘the central part of the present dispute [was] the application of income tax laws which is covered under the DTT’ and thus the claims fell ‘within the ambit of the mutual agreement procedure’. ¹³¹ More concretely, Poland pointed out that the dispute was covered under Articles 8, 11, 13, and 15 of the DTT. ¹³²

The Tribunal noted that ‘the central issue in this case relates to deductibility of [management] costs for the purposes of calculating corporate income tax and the DTT is applicable to income tax’. ¹³³ Despite this, after examining the DTT provisions relied upon by Poland, the Tribunal found that the dispute was not subject to the MAP. With respect to Article 8 of the DTT, the Tribunal observed that this provision ‘deals with business profits and sub-article (3) provides that in determining the profits of a business, deductions for expenses incurred for the purposes of the business shall be allowed’. ¹³⁴ The Tribunal held that this provision was not relevant in that case at hand since, according to the ‘Respondent’s own formulation’, the dispute was ‘not whether Management Services were in fact provided, but whether [the claimants] adequately documented the provision of the Management Services for the purposes of claiming deductions’. ¹³⁵ With respect to Article 13 of the DTT, which relates to royalties, the Tribunal found that ‘there is no issue of royalties being paid to or by anyone in this case and therefore, Article 13 of the DTT has no application to the present dispute’. ¹³⁶ As to the Tribunal’s position on Article 15 of the DTT, which deals with the taxation of income derived from the provision of services, such as management services, the Tribunal found the dispute did not concern this issue but rather ‘the treatment of the expenses incurred by [the claimants] in paying for the Management Services’. ¹³⁷

Finally, the Tribunal decided that ‘the Claimants’ claim that their freedom to transfer funds was violated because they could not freely transfer the Management Fees [was] very different from the taxation of dividends covered under Article 11’ and thus that provision was not applicable either. ¹³⁸

Based on the above analysis, the Tribunal held that it only had jurisdiction to hear the claimants’ claims based on expropriation and transfers of funds pursuant to Article VI(2) of the BIT. On the merits, however, the Tribunal decided that both claims had failed, and the claimants were, consequently, not entitled to any damages. ¹³⁹

These three cases give a glimpse of how the investment and tax treaty regimes have the potential to interact and overlap. With the support of arbitral jurisprudence, and in the absence of a clear definition regarding the relationship between IIAs and DTTs, disputes arising from tax related measures can fall within the scope of both treaty regimes. This can lead to the parallel use of the dispute settlement mechanisms offered in each field to resolve disputes arising out of the same measure. As UNCTAD explains:

Potentially, a taxpayer could request the relevant competent authority for a mutual agreement procedure (MAP) and, concurrently or afterwards, pursue ISDS claims as an investor under an IIA concerning the same matter. A MAP between the competent authorities of the contracting parties or a State-State tax arbitration could be ongoing when an ISDS proceeding is initiated. The outcome of a MAP, tax arbitration or tax litigation could also give rise to ISDS cases. ¹⁴⁰

In other words, the proliferation of overlapping and unco-ordinated mechanisms to resolve tax disputes in the international plane have resulted in further fragmentation. This demonstrates, as Chaisse aptly observes, that ‘there is a need for better designed international rules and policies on tax and investment, which would allow the tax and investment worlds to move from mere coexistence to cooperation’. ¹⁴¹

The opportunity given to investors to challenge tax policies before ISDS tribunals ‘also create an uncertain environment for states considering tax legislation, potentially chilling the development of their fiscal regime’. ¹⁴² An illustrative example is the decision taken by the Indian government to withdraw its retroactive tax demands in order to settle the Cairn v. India and related arbitrations. ¹⁴³ In the ConocoPhillips v. Vietnam case, however, the parties reached a settlement by which, according to Finance Uncovered, ConocoPhillips finally agreed to pay capital gain taxes on the sale of its two subsidiaries. This can set a precedent for other states that wish to levy taxes to fund public policy objectives. As a journalist put it when discussing the oil industry’s ‘pre-emptive legal strike’ on Vietnam: ‘As more countries claim their resources have been bought and sold by foreigners tax-free, this issue is likely to become a new frontier in the anti-tax avoidance campaign.’ ¹⁴⁴

States are also responding to the increasing volume of ISDS claims involving tax measures by amending investment treaty provisions. Preserving tax policy autonomy and co-ordinating investment and tax disputes settlement mechanisms has become an important matter for treaty negotiators. As compared to old-generation IIAs, more recent treaties contain tax carve-out provisions that aim to completely exclude tax measures from their scope of application and to avoid overlap between IIAs and the subject matters covered by DTTs. ¹⁴⁵ The 2016 India Model, for instance, states that the treaty shall not apply to ‘any law or measure regarding taxation, including measures taken to enforce taxation obligations’. ¹⁴⁶ This article further provides that a host state’s decision that a particular regulatory measure is related to taxation, whether made before or after the commencement of arbitral proceedings, ‘shall be non-justiciable and it shall not be open to any arbitration tribunal to review such decision’. ¹⁴⁷ As Ranjan notes, ‘it is evident that India has decided to keep taxation measures outside the purview of the BIT in response to Vodafone and Cairn challenging India’s retrospective application of taxation law under different BITs’. ¹⁴⁸

The new 2018 Dutch model BIT mostly aims at avoiding conflicts with DTTs. Article 10(3) provides that:

[t]his Agreement does not affect the rights and obligations of a Party under an agreement for the avoidance of double taxation. In the event of inconsistency between such agreement and this Agreement, the agreement for the avoidance of double taxation prevails to the extent of the inconsistency. ¹⁴⁹

This clause, however, does not clarify how and by whom (ISDS tribunal or domestic tax authorities) inconsistencies should be settled. This means that, as occurred in Schooner v. Poland, it would be for the ISDS tribunal to determine whether there is a conflict between the IIA and the DTT. The position of the tribunal regarding a potential conflict between the two regimes may be different from that adopted under domestic law.

Only a few IIAs, such as Article 14(4) of the Chile-Hong Kong BIT, contain a provision specifying that any determination as to the existence of an inconsistency between a DTT and an IIA shall be settled by the competent (tax) authorities of the contracting parties. ¹⁵⁰ Under this mechanism, the designated competent authorities shall take a decision ‘within six months of the referral of the issue’ and such a decision is binding on the arbitral tribunal. ¹⁵¹ However, ‘[i]f the designated authorities have not determined the issue within six months from the date of the referral, the tribunal or arbitral panel shall decide the issue’. ¹⁵²

The Chile-Hong Kong BIT indeed gives a greater role to host states on taxation claims. This treaty, however, does not address a possible scenario where the competent tax authorities do not reach an agreement within the deadline established in the treaty, leaving that decision to the tribunal. That outcome may differ from the one of the competent tax authorities if the same claim is brought under a DTT.

Putting aside these drafting deficiencies, these treaties only represent the minority of the vast IIA universe. As previously explained, most IIAs do not address tax issues and are silent on the relationship between their scope of protection and the subject matter covered by DTTs and MAPs. In the absence of a consistent and coherent treatment of fiscal matters in tax and investment treaties, investors will continue to challenge tax policies before ISDS tribunals that may potentially fall within the realm of DTTs. States will, in turn, continue to face the consequences of regulatory gaps, such as unintended and expansive interpretations of treaty provisions. It is thus necessary to establish a more effective safeguard for avoiding overlaps between IIAs and tax policymaking. Recent reform efforts taking place mostly at the bilateral level show how states are trying to achieve this goal.

Amending or renegotiating IIAs bilaterally or even regionally may not, however, be the most effective way to advance harmonization between the regimes. This approach will result in continued fragmentation and will foster treaty shopping by investors that restructure their investments through companies incorporated in states that have signed IIAs that do not contain tax carve-out provisions. It is submitted that any reform efforts designed to improve coherence between IIAs and DTTs will first require the co-operation between investment and tax policymakers. They should avoid the formulation of investment and tax policymaking in vacuums. In this respect, they should seek to minimize the risk of friction between the existing unco-ordinated and potentially overlapping dispute resolution mechanisms established in investment and tax treaties. This can be achieved through the establishment of a standing multilateral mechanism.

5. Concluding remarks

An essential objective of any dispute resolution system should be to provide predictability and certainty and to consistently deliver comprehensive justice. It is doubtful whether the different mechanisms implemented in the tax and investment law regimes will achieve that objective, at least individually. This article has shown that the fragmentation of investment and tax dispute settlement may not the way forward to address the different concerns pertaining to each regime. In its current form, the different and unco-ordinated mechanisms established in the tax regime fail to effectively accommodate the interests of taxpayers on the presumption that this will ensure respect for states’ sovereignty in international fiscal policy. Further, these mechanisms can lead to inconsistency and unpredictability in the interpretation of tax treaties. In investment law, the different and unco-ordinated systems established in IIAs leave a number of concerns unresolved, including the asymmetry prevalent in the investment treaty regime and inconsistency in the interpretation of treaty provisions. The horizontal structure of investment and tax dispute settlement is also evidencing the side effects of fragmentation. The largely undefined interplay between ISDS and MAP can undermine the rights and obligations secured in DTTs and allow investors to evade tax policies.

A fruitful alternative to fragmentation would be to establish a centralized multilateral standing mechanism for the settlement of investment and tax disputes. Current efforts to create that mechanism for the international investment regime are ongoing under the auspices of the UNCITRAL Working Group III. Proposals to address the concerns relating to ISDS include the establishment of a Multilateral Investment Court (MIC). ¹⁵³ Depending on its design, a MIC has the potential to offer a unified and more coherent framework through which ISDS shortcomings can be addressed in a collective manner. As Schill and Vidigal observe, a MIC ‘would be capable of addressing most, if not all, the concerns identified in the UNCITRAL process: it would ensure greater coherence and consistency and be more independent and cost-effective than any alternative’. ¹⁵⁴

If implemented with a permanent two-tier adjudicative system, as proposed by the EU and other states, ¹⁵⁵ a MIC would ensure predictability and consistency in the interpretation of treaty provisions. As Luca notes, ‘[c]concentrating interpretative authority in one centralized, multilateral court – rather than distributing it among several appellate bodies that might be attached to particular IIAs – could greatly advance predictability on a systemic level with respect to the application of IIA provisions’. ¹⁵⁶ It would further ‘create opportunities for providing authoritative interpretive guidance on the correct identification and precise application of applicable law under IIAs’. ¹⁵⁷ Moreover, if composed of a permanent body of tenured (or semi-tenured) judges, instead of party-appointed arbitrators, the MIC could reduce the problems associated with independence and impartiality as it would weaken the link between adjudicators and counsel for investors and states. The Working Group has also proposed to include in the MIC ‘a mechanism to cater for possible counterclaims by respondents’, allowing states to bring claims against investors for a breach of obligations under an investment treaty. ¹⁵⁸ This would mitigate unbalanced provisions prevalent in most IIAs

However, certain questions relating to the creation of a MIC remain open and must be addressed. The Working Group III has clarified that a MIC ‘would adjudicate over the relevant underlying international investment instruments, rather than one sole investment treaty with a unified set of substantive standards and provisions’. ¹⁵⁹ A key challenge for the MIC, therefore, would be to harmonize textual differences across substantive standards IIAs. As Bjorklund highlights:

[o]ne of the hopes for the MIC is to bring greater consistency and coherence to investment law. How it can do this when the idea is that more than 3000 treaties could feed into one court—and especially how it can do this when those treaties have different provisions in them—is a bit unclear. ¹⁶⁰

Another question relates to the willingness of states to agree on the implementation of a MIC that will resolve investment disputes arising out of their IIAs. Some authors take an optimistic stance in this regard, arguing that:

the acceptance by the CJEU of the EU’s participation in international mechanisms for the adjudication of investment disputes, combined with the EU’s efforts towards establishing a MIC, greatly increase the likelihood that some form of multilateral institution will emerge over the coming years. ¹⁶¹

Lastly, the Working Group III should discuss how the MIC should deal with investment disputes relating to tax measures. The development of a MIC could certainly be considered as an option in the tax law regime, instead of the fragmented approach to dispute settlement adopted through the MIL and the EU Dispute Resolution Directive. Indeed, as has been argued, ‘the uniformity of tax treaty interpretation can be better safeguarded and further developed when only one or two international courts are involved, than in the case of a myriad of bilateral arbitration tribunals’. ¹⁶² Creating a centralized system of dispute resolution for tax treaty disputes will not, however, prevent investors from challenging tax policies before ISDS tribunals. It is at this point that co-operation between tax and investment policymakers becomes crucial. Ongoing discussions on the MIC present a good opportunity to establish this co-operation. A key item for discussion should be how to create more certainty regarding the relationship between international rules and policies on tax and investment and, in particular, how the MIC should address situations where a tax-related matter may simultaneously fall within the scope of a DTT as well as an IIA between the relevant countries.

This article has offered an initial mapping of the problems created by the fragmentation of unco-ordinated international and tax dispute settlement mechanisms. This field of inquiry is in its infancy, but the issues for policymakers are pervasive.