CONTROL MECHANISMS IN MULTILATERAL INVESTMENT TRIBUNALS: NAVIGATING PROCEDURAL MULTILATERALISM AND SUBSTANTIVE BILATERALISM

Abstract

A future multilateral investment court (MIC) or multilateral appellate mechanism (MAM) will operate on a plurilateral basis, among States that become parties to the tribunal's constitutive instrument and grant it jurisdiction over disputes under their investment treaties. The creation of a MIC or MAM would involve a significant strengthening and centralization of dispute settlement institutions in the investment treaty regime, which is already overly dependent on law-development by adjudicators, reflected in well-established concerns about loss of State control. Thus, a key challenge in designing a MIC or MAM is to incorporate appropriate control mechanisms that will enable State input, without unduly undermining a MIC or MAM's independence. This article analyses control mechanisms in a MIC or MAM, considering a wide range of questions of institutional design. It highlights two fundamental tensions. One is the tension between independence and accountability. The other tension is between procedural multilateralism and substantive bilateralism. While the procedural law in a MIC or MAM will have been multilateralized, the substantive law the tribunal will interpret and apply will remain contained in mostly bilateral investment treaties, controlled by the parties to those agreements. This article addresses the challenges of designing a multilateral tribunal for a regime that lacks multilateral substantive law and contributes to wider debates over striking an appropriate balance between international judicial independence and Member State control.

I. INTRODUCTION

A multilateral investment court (MIC) or multilateral appellate mechanism (MAM) for investment disputes is no longer an abstract or academic proposition. Draft provisions on the design of a standing, two-tier multilateral mechanism for the resolution of investment disputes (referred to in this article as a MIC)—which may include as one component a standing appellate tribunal that States can opt into without accepting the first-instance tribunal (referred to as a MAM)—have been discussed at the United Nations Commission on International Trade Law (UNCITRAL) Working Group III in recent years and remain on the Working Group's agenda.¹ Both a MIC and a MAM are potential outcomes of the Working Group III process and both are likely to operate on a plurilateral opt-in basis as they are only supported by certain States. Specifically, it has been clear since the early days of Working Group III that there is a subset of States—most notably the European Union (EU) and its Member States—that favour the creation of a two-tier MIC,² and another, potentially broader, group of States, including China and Morocco, which favour a standing MAM to sit above the existing system of ad hoc arbitration.³ As noted above, the appellate function could be performed by the appellate tier of a MIC if a so-called ‘open architecture’ is pursued whereby States can opt in to a MIC's appellate jurisdiction only.⁴ While a MIC and a MAM are distinct reform proposals, this article treats both options together because they raise common questions in relation to the themes that are the focus of this article: ensuring State control and designing a multilateral adjudicatory institution for a regime where the substantive law will remain largely bilateral.

The creation of a MIC or MAM (or both) would constitute a significant strengthening and centralization of dispute settlement mechanisms in the investment treaty regime.⁵ Yet, as Alschner has noted, the investment treaty regime is already overly dependent on law-development by arbitral tribunals, reflected in well-established concerns about loss of State control.⁶ Accordingly, a further strengthening and centralization of adjudication, without an associated strengthening of mechanisms for State input and control, is unlikely to prove sustainable.⁷ Thus, as Yu notes in passing, if States choose to create a MIC or MAM (or both), steps should be taken in the design of the tribunal to address the significant risks of backlash arising from the new centralized tribunal's increased law-making influence.⁸

This article considers the design of appropriate control mechanisms in a MIC or a MAM. In a nutshell, a MIC or MAM needs to include control mechanisms that provide adequate opportunities for State input and control over the tribunal and its influence on the investment treaty regime. It is only with adequate avenues for State input and control that there may be long-term political buy-in to the jurisdiction of a MIC or MAM.⁹ However, a MIC or MAM will also require substantial independence from the States Parties that create it if it is to fulfil the functions expected of it, such as interpreting international investment agreements (IIAs) in a principled manner which takes account of both State and investor interests.¹⁰ Thus, as Roberts and St John argue, one ‘polarity’ or tension that must be ‘managed’ on an ongoing basis by policymakers in investor–state dispute settlement (ISDS) reform processes—and particularly in designing a MIC or MAM—is a ‘tension between independence and accountability’.¹¹ It will be shown throughout this article that the tension between independence and accountability underlies numerous questions concerning the design of a MIC or MAM. While a MIC or MAM should be designed to include adequate mechanisms for State control, to ensure responsiveness to Member States and long-term political buy-in to the institution, it is also important that such control mechanisms are designed with safeguards against their abuse in mind. Although there is an existing literature on control mechanisms in the investment treaty regime, as Roberts and St John note, this largely focuses on the exercise of control mechanisms on a unilateral or bilateral basis, and how States can effectively exercise control in a multilateral setting is less well understood.¹²

In considering the design of control mechanisms for a MIC or a MAM, this article gives particular attention to issues arising from the fact that while the procedural law in a MIC or MAM will have been multilateralized in the Statute of the tribunal and its procedural rules, the substantive law that the tribunal will apply will be contained in mostly bilateral investment treaties (BITs). At one level this simply reflects the procedural focus of the UNCITRAL Working Group III process, whose mandate intentionally focuses on reform of ISDS procedures rather than on the substantive standards in IIAs.¹³ Furthermore, as Titi notes, the phenomenon of procedural multilateralism—despite the substantive law remaining bilateral—is not entirely new in investment law, as illustrated by procedurally focused multilateral instruments such as the Convention on the Settlement of Investment Disputes between States and Nationals of Other States (ICSID Convention) and the United Nations (UN) Convention on Transparency in Treaty-based Investor–State Arbitration.¹⁴ Yet such an approach means that certain control mechanisms will remain solely in the hands of the parties to the relevant IIA—for example, there is little doubt that adopting authoritative interpretations of an IIA, treaty amendments, or terminating an IIA will remain control mechanisms exercisable exclusively by the parties to the relevant IIA.¹⁵ In contrast, other control mechanisms are likely to involve all States Parties to the Statute of a MIC or MAM—for example, electing the members of the tribunal, or adopting the tribunal's budget. Third-party involvement in proceedings before a MIC or MAM is an area where this tension between procedural multilateralism and substantive bilateralism will be particularly acute. For example, should States that are parties to the Statute of a MIC or MAM, but not parties to the particular IIA invoked in a dispute, be given some ability to make submissions in disputes that raise issues concerning interpretation of the tribunal's Statute, or that raise issues of wider systemic importance for numerous other IIAs? Given one of the aims of creating a MIC or MAM is for the standing tribunal to provide greater consistency in the interpretative approach applied to IIAs,¹⁶ it is foreseeable that initial positions taken by a MIC or MAM, for example regarding terms or concepts contained in numerous IIAs, will prove influential in later cases concerning other IIAs.¹⁷

This article's contribution lies not only in its analysis of a wide range of questions of institutional design of a MIC or MAM, but also in its elaboration of a conceptual framework for understanding two fundamental tensions that underlie and animate many of the specific questions of institutional design. These two recurrent tensions are those introduced above: the tension between independence and accountability or State control; and the tension between procedural multilateralism and substantive bilateralism. Through its analysis of control mechanisms in a future MIC or MAM, this article contributes to broader debates in international (economic) law of how to balance the independence of international tribunals with control by, and responsiveness to the preferences of, Member States.¹⁸

This article proceeds as follows. Section II analyses several key concepts on which the article relies, namely mechanisms for State control in international tribunals and the concepts of independence and accountability of international tribunals. Section III considers those control mechanisms that are likely to operate on a multilateral basis, involving all States Parties to a MIC or MAM. Section IV addresses control mechanisms that are likely to remain solely in the hands of the parties to the particular IIA at issue. Section V then turns to control mechanisms that raise the tension between procedural multilateralism and the substantive law that a MIC or MAM will apply being largely contained in bilateral treaties controlled by the parties to those agreements. Among other issues, Section V considers approaches to third-State participation in proceedings before a MIC or MAM. Section VI concludes.

II. DESIGNING CONTROL MECHANISMS

Thinking about the design of a MIC or MAM does not start from a blank slate. Rather, it can rely on a significant existing literature and experience in other contexts concerning mechanisms for State control in international tribunals.¹⁹ This section reviews existing knowledge and experience regarding control mechanisms in international tribunals, which can inform debates about the design of a MIC or MAM. It also reviews the concepts of judicial independence and judicial accountability, as these are key values which must be balanced in designing control mechanisms.

A. Mechanisms for State Control in International Tribunals

‘Control’ has been defined in this context as ‘checks on the powers of an [international] organization that ensure that the organization acts within its assigned mandate’.²⁰ This subsection does not purport to introduce every possible control mechanism, but to introduce those most relevant to a MIC or MAM, which will then be analysed in subsequent sections.²¹ At the outset, it is worth noting one common consideration that applies to most of the control mechanisms discussed below, which is directly relevant to ongoing policy debates over designing a MIC or MAM. In short, because many control mechanisms require States to act as a collective principal—ie to coordinate and reach agreement, which is often difficult, particularly in a multilateral context due to States’ diverging preferences—the various means of State control over international tribunals are less effective than might be anticipated.²²

One well-recognized control mechanism is precision in drafting the substantive law that a tribunal applies.²³ As an ex ante control mechanism this tool is familiar in the investment treaty regime, in the form of States’ attempts to draft more precise IIAs over the last 20 years.²⁴ In the investment treaty regime, adding greater precision ex post takes the form of authoritative interpretations agreed by treaty parties, agreements to amend the relevant IIA, or agreements to negotiate an entirely new IIA to replace the existing treaty. A related control mechanism known in other international tribunals is specifying a particular interpretative approach that the tribunal must follow.²⁵ A well-known example is Article 3.2 of the World Trade Organization's (WTO) Dispute Settlement Understanding (DSU), which provides that one function of WTO dispute settlement is to clarify the provisions of WTO-covered agreements ‘in accordance with customary rules of interpretation of public international law’ and that ‘[r]ecommendations and rulings of the DSB [Dispute Settlement Body] cannot add to or diminish the rights and obligations’ of members.²⁶ Other precedents also exist, for example the deferential standards of review adjudicators are instructed to apply under Article 17.6 of the WTO Anti-Dumping Agreement,²⁷ or the instruction in the WTO Multi-Party Interim Appeal Arrangement (MPIA) that ‘[t]he arbitrators shall only address those issues that are necessary for the resolution of the dispute. They shall address only those issues that have been raised by the parties.’²⁸ While specifying a particular interpretative approach can have an important signalling effect for adjudicators, it is difficult for treaty negotiators through such ex ante formal legal regulation to control precisely how a tribunal will perform its functions, including as tribunals are typically left to interpret their own mandate.²⁹

Another important category of control mechanisms concerns the appointment of members of an international tribunal. At the stage of drafting a constitutive instrument, choices facing treaty negotiators include the term length for appointments, whether to provide for renewable terms, which bodies elect judges and by what voting rules, the criteria specified for judges in terms of required expertise, geographic diversity and other forms of diversity, and whether a process for screening nominees for suitability prior to appointment is created.³⁰ At the stage of drafting a constitutive instrument, the States Parties to a tribunal exercise these control mechanisms collectively, as it is only with the agreement (or at least non-objection) of all treaty parties that such features will end up in a constitutive instrument. In contrast, at the stage of making (re)appointments to a tribunal, there may be some scope for control to be exercised by individual States parties, or a subset of States parties to the tribunal, for example if there is a requirement that candidates are nominated by a Member State of the tribunal or by a regional grouping, or if powerful States are able to exercise a veto.³¹ The last point highlights that the voting rules applicable to a particular decision are a crucial element of institutional design and will affect whether a particular control mechanism operates collectively, involving all States Parties to a tribunal, or gives rise to some possibility for unilateral control by individual Member States or control by a subset of Member States. For example, if a particular decision (eg appointment of tribunal members) is made subject to a requirement of consensus, defined as no Member State objecting, this will mean that a control mechanism that operates multilaterally in practice gives rise to a significant risk of unilateral vetoes by individual Member States, and particularly by powerful Member States which are able to wield a threat of veto credibly.³² Interestingly, the UNCITRAL secretariat Draft Statute of a Standing Mechanism provides that where the Conference of the Contracting Parties—the plenary body of Member States—cannot make a decision by consensus, decisions may be made by a four-fifths majority of Contracting Parties present and voting.³³ This approach, if adopted, would eliminate the risk of unilateral vetoes and substantially reduce the risk of the tribunal being held hostage by a small minority of Contracting Parties.

A third important mechanism for State control is a tribunal's budget.³⁴ As Cogan notes, international ‘courts are entirely dependent on States and international organizations for their funding’.³⁵ Most international tribunals are primarily funded from the assessed budgetary contributions of the Member States—either the Member States of the relevant international tribunal (as in the case of the International Criminal Court (ICC)) or the Member States of the parent international organization where the tribunal is funded from the budget of a parent organization (as in the case of the International Court of Justice (ICJ) and the WTO Appellate Body (AB)).³⁶ Typically the budget must be adopted by the plenary political body that oversees the relevant international tribunal, eg the UN General Assembly for the ICJ, or the Assembly of States Parties (ASP) for the ICC.³⁷ While a tribunal's budget may formally need to be approved by the plenary body of States that governs a tribunal, the issue of budgetary control also highlights that certain control mechanisms are in practice more available to some Member States than others.³⁸ For example, a threat to withhold paying assessed budgetary contributions is obviously far more serious for a tribunal when it comes from a major donor.³⁹

A fourth category of control mechanisms concerns the design of rules concerning the jurisdiction of a tribunal.⁴⁰ Most obviously this concerns which kinds of actors are given access to a tribunal as parties, whether an additional act of State consent is needed for the tribunal to have jurisdiction besides the constitutive instrument creating the tribunal, and which kinds of disputes a tribunal is given jurisdiction over (eg what subject matter limitations are imposed). Carve-outs of sensitive issues from any advance grant of State consent to adjudication are an important control mechanism that treaty drafters enjoy.⁴¹ It is widely recognized that tribunals that confer standing on non-State actors are much harder for States to control than tribunals that only permit State–State claims, given the likelihood of a far higher volume of cases in the former scenario and the loss of State control over which cases are brought before a tribunal.⁴²

A fifth broad category of control mechanism concerns the procedural law to be applied by a tribunal. While many constitutive instruments delegate to an international tribunal the power to determine its own procedures,⁴³ alternatives exist granting States greater control. An example is the Rome Statute of the ICC, which provides that the Rules of Procedure and Evidence (and amendments thereto) must be adopted by a two-thirds majority of members of the ASP.⁴⁴ While there is provision for the ICC judges, by two-thirds majority, to adopt provisional rules where the Rules of Procedure ‘do not provide for a specific situation before the Court’, these can be adopted, amended or rejected by the ASP.⁴⁵ Although the Rome Statute provides that the ICC judges shall adopt ‘the Regulations of the Court necessary for its routine functioning’ by majority and these take effect upon adoption, they must be circulated to States Parties for comments and only remain in force if there is not an objection within six months from a majority of States Parties.⁴⁶

A sixth potential control mechanism lies in the provisions governing the secretariat of an international tribunal. At one level, a secretariat can be a means of controlling adjudicators and the potential for rogue interpretations.⁴⁷ However, significant concerns have been raised about the oversized and unaccountable influence of secretariats, particularly in WTO dispute settlement.⁴⁸ Such concerns have clearly influenced thinking about the design of a secretariat for a MIC or MAM, for example the provision in the current Draft Statute for term limits for the Executive Director of the Secretariat and clarification that the Executive Director ‘shall be accountable to the Conference [of the Contracting Parties]’.⁴⁹

A seventh group of control mechanisms concerns the kinds of remedies a tribunal is authorized to award, the legal status of decisions rendered, and the practical scope for non-compliance by States.⁵⁰ Constitutive instruments can limit the scope of remedies open to a tribunal. One example is the system of remedies under the WTO DSU, which removes the potential for an award of compensation to redress past wrongful conduct.⁵¹ Another is the common specifications in newer investment treaties that if a tribunal orders restitution of property, it must provide an alternative to the State of paying compensation in lieu of restitution, and that a tribunal may not award punitive damages.⁵² While most constitutive instruments provide that a tribunal's decisions are binding for the disputing parties as a matter of international law, there are important variations regarding whether tribunals’ decisions are made enforceable in domestic legal systems.⁵³ For present purposes, the key point is that in settings where judgments of international tribunals are enforceable in domestic legal systems, this reduces State control over the system.⁵⁴ The existing system of investor–State arbitration sits towards the strong end of the remedies/enforcement spectrum, as tribunals’ decisions are enforceable in domestic systems pursuant to powerful multilateral frameworks, principally the ICSID Convention and the UN Convention on the Recognition and Enforcement of Foreign Arbitral Awards.⁵⁵

B. Independence and Accountability of International Tribunals

In the context of international tribunals, judicial independence is typically understood as a ‘set of institutional and other factors that, to a lesser or greater extent … allows judges to develop legal opinions unconstrained by the preferences of other actors’.⁵⁶ Existing discussions largely focus on the independence of international tribunals from the States that create and maintain such institutions.⁵⁷ For example, Keohane, Moravcsik and Slaughter defined independence as ‘the extent to which adjudicators … are able to deliberate and reach legal judgments independently of national governments’,⁵⁸ arguing:

[t]he extent to which members of an international tribunal are independent reflects the extent to which they can free themselves from at least three categories of institutional constraint: selection and tenure, legal discretion, and control over material and human resources.⁵⁹

It is important to remember that judicial independence is not an absolute value, but is balanced against other competing interests,⁶⁰ which include responsiveness to the States that create and sustain an international tribunal.⁶¹

Judicial accountability is ‘in many ways merely the other side of the coin to judicial independence’.⁶² Accountability is often understood as the principle whereby ‘some actors have the right to hold other actors to a set of standards, to judge whether they have fulfilled their responsibilities in light of these standards, and to impose sanctions if they determine that these responsibilities have not been met’.⁶³ Wong and Yackee emphasize that the point of the sanctioning involved in accountability is ultimately to influence the actor being held to account and to incentivize that actor to act in future consistently with the preferences of the actor doing the holding to account.⁶⁴ While a variety of actors could have a role in holding international tribunals accountable,⁶⁵ in terms of control mechanisms what is of interest is accountability to the States that create and sustain an international tribunal.⁶⁶ As noted above, it is possible to distinguish between control mechanisms that are exercised collectively by all parties to a tribunal's constitutive instrument, and control mechanisms that may be open to individual States Parties, or a subset of States Parties to the constitutive instrument.⁶⁷

III. MULTILATERAL CONTROL MECHANISMS

This section considers those control mechanisms that will likely be exercised by all States Parties to the statute of a MIC or MAM. The UNCITRAL secretariat Draft Statute of a Standing Mechanism includes the creation of a plenary body, ‘the Conference of the Contracting Parties’, which is composed of representatives of all Parties to the instrument establishing the Standing Mechanism, and is given broad responsibility to oversee the functioning of the Standing Mechanism.⁶⁸ The creation of such a plenary body follows the example of other ‘international judicial governance institutions’, for example the DSB in relation to WTO panels and the AB, the ASP in relation to the ICC, or the Meeting of States Parties to the UN Convention on the Law of the Sea (SPLOS) in relation to the International Tribunal for the Law of the Sea (ITLOS).⁶⁹

A. Appointment of MIC or MAM Members

It is likely that all parties to a MIC or MAM will collectively exercise control over the appointment of members to the tribunal or tribunals. However, as explained above, much will depend upon the decision-making rules that apply.⁷⁰ The UNCITRAL secretariat Draft Statute of a Standing Mechanism gives the Conference of the Contracting Parties—the plenary body of all Member States—the role of appointing the members of the first-instance tribunal and appeals tribunal,⁷¹ and of establishing a Selection Committee that has the role of screening nominees to determine whether they meet the qualification requirements, prior to the Conference making appointments to the tribunals.⁷² From the perspective of State control, it is notable that within the Working Group III process there appears to be substantial support for a screening mechanism, with the purpose of screening out candidates that do not meet the relevant criteria, in between the nomination of candidates and their appointment.⁷³ This approach is clearly inspired by the screening mechanisms developed in recent years in certain other international tribunals.⁷⁴ Such a screening mechanism offers reduced State control compared to a system where States have complete discretion to judge the suitability of candidates. However, State control is not the only relevant value, and it may be that the parties to a MIC or MAM are willing to relinquish some discretion if it provides a greater guarantee of only qualified candidates proceeding to the election stage. Similarly, it is noteworthy that, following earlier State feedback, the UNCITRAL secretariat Draft Statute permits the Conference of the Contracting Parties to make an open call for candidates—ie to expand the range of actors that may nominate candidates beyond Member States.⁷⁵ Again, this would represent a weakening of State control—compared to a system only permitting nominations by Member States—in the interest of furthering other values (eg a transparent and depoliticized process for judicial appointments).⁷⁶

Another interesting design choice concerns whether all parties to the instrument establishing a MIC or MAM will be entitled to vote on the election of members, or only those States that accept the jurisdiction of the relevant tribunal. This will be a particular issue if an ‘open architecture’ is pursued whereby States can opt into different parts of the standing tribunal (eg only the appellate tribunal).⁷⁷ The question will then be, do States that have only accepted the appellate tribunal's jurisdiction have a right to participate in decisions concerning the first-instance level of the MIC (eg election of first-instance tribunal members)? An earlier UNCITRAL secretariat draft appeared to suggest that accepting the jurisdiction of a MIC need not be a requirement for parties to the MIC's statute to vote on the election of judges, noting ‘there are several courts in which tribunal members are selected by treaty parties or by a collective body of States, even if that membership is larger than the group of States that accept the court's jurisdiction’, and citing examples such as the ICJ and ITLOS.⁷⁸ In comments on this draft, the EU objected to the possibility of members of the MIC being elected by States ‘other than those that accept the tribunal's jurisdiction’.⁷⁹ The most recent secretariat documents acknowledge that this issue remains to be resolved.⁸⁰ If the ability to vote on electing members of a MIC or MAM is made conditional on a State's acceptance of the jurisdiction of the relevant tribunal, one question that will arise is what degree of acceptance of jurisdiction is required? For example, would a State's acceptance of the jurisdiction of the appellate tribunal for a single dispute, or for disputes under just one of its IIAs, mean that the State is entitled to vote on appellate tribunal appointments?

Staying with the election of members to the MIC or MAM, another interesting and yet-to-be-resolved design choice is whether all State Parties to the MIC or MAM would be entitled to vote on all nominated candidates, or whether the various regional groupings—seen as important to ensure geographic diversity—would only vote on candidates from their region.⁸¹ While an initial secretariat draft appeared to suggest the latter approach,⁸² in discussions in Working Group III ‘it was generally felt that the members of the Committee of the Parties should be entitled to vote on all identified candidates and not limited to those that fall within their regional group’,⁸³ and several States in written comments have opposed the idea of States only voting on candidates from their regional groups.⁸⁴ If voting is not confined to regional groups, then the control mechanism of deciding on judicial appointments will operate more diffusely, involving all States Parties to a MIC or MAM, assuming that there are more nominees from each regional group than seats allocated to the group.⁸⁵

B. Setting the Budget for a MIC or MAM

Another control mechanism that will likely be exercised by all States Parties to a MIC or MAM concerns the tribunal's budget. Given that only a subset of States support the reform option of creating a MIC or a MAM, it is unlikely that either body could be funded from the budget of another international organization (eg the UN budget).⁸⁶ Rather, a MIC or MAM is likely to be funded from contributions of the States Parties to these bodies, similar to the ICC (funded from assessed contributions of the States Parties to the Rome Statute)⁸⁷ or ITLOS (funded by the States Parties to the UN Convention on the Law of the Sea and the International Seabed Authority).⁸⁸ This could make the MIC or MAM relatively vulnerable to control in the form of funding cuts.

In the UNCITRAL secretariat Draft Statute, the Conference of the Contracting Parties is given broad control over the Standing Mechanism's funding. For example, the Conference is to ‘[a]dopt the annual budget of the Standing Mechanism’ and to ‘[d]etermine the amount of remuneration of the members’ of the first-instance and appeals tribunals.⁸⁹ The obvious challenge here will be to design the funding arrangements so that fixing the annual budget, including fixing the compensation of tribunal members, does not become a mechanism of political control by States Parties.⁹⁰ Reflecting this concern, one of the overarching general principles to govern the operation of the Standing Mechanism is that it shall be ‘independent and free from undue external influence, including from its donors’.⁹¹ An interesting precedent that could insulate a MIC or MAM from Member State pressure via a reduction of funding is the model of the Caribbean Court of Justice, where the Court is funded through the income earned on an independently managed trust fund, which was established by the States that established the Court to provide for the Court's running costs (including salaries).⁹² As Malleson has argued, this funding model may have ‘important benefits in securing the long-term independence of … [international] courts’ and ‘warrant[s] serious consideration for wider adoption by the community of international courts’.⁹³

A MIC or MAM may also be partly funded through the charging of user fees, similar to the administrative fees that disputing parties are charged by arbitral institutions.⁹⁴ Indeed, the UNCITRAL secretariat Draft Statute foresees the Conference of the Contracting Parties adopting a fee structure for the Standing Mechanism.⁹⁵ Obviously, charging user fees could reduce the MIC or MAM's dependence on Member State contributions. However, there is a clear concern in the Working Group III process—given the issues in the existing ISDS system that have motivated the reform process—to avoid any direct link between user fees, the volume or length of cases, and remuneration of members of a MIC or MAM.⁹⁶ Thus the UNCITRAL secretariat has noted that:

[i]t may be prudent that the budget for the operation of the Tribunals (in particular, the remuneration of the members of the Tribunals) relies only on contributions by Contracting Parties rather than on fees to be charged … [to] ensure the independence and integrity of the Tribunals.⁹⁷

C. Control over Procedural Law to be Applied by a MIC or MAM

Another control mechanism likely to be exercised by all States Parties to a MIC or MAM concerns the procedural law to be applied by the tribunal. As noted above, existing international tribunals demonstrate there are a variety of models, from delegating the creation of rules of procedure to the tribunal itself, to tribunals such as the ICC where Member States retain far greater control.⁹⁸ The UNCITRAL secretariat Draft Statute gives the Committee of the Contracting Parties the power to adopt the rules of procedure of the first-instance and appellate tribunals, and to ‘[a]dopt administrative, financial and other regulations on the operation of the Standing Mechanism’.⁹⁹ These provisions suggest that States Parties to a MIC or MAM are keen to retain control over the procedural law to be applied by the new tribunal. However, some form of residual power over procedures will need to be afforded to a MIC or MAM, given that the rules of procedure and other regulations adopted by the Committee of the Contracting Parties are unlikely to provide for every scenario that may arise. For example, in discussion of these aspects of an earlier UNCITRAL secretariat draft, ‘it was suggested that flexibility should be given to the Tribunal to update its rules and adapt its procedure when necessary’,¹⁰⁰ and several States highlighted the need for further thought regarding how to divide the power over procedures between the Tribunal and Member States.¹⁰¹ As noted above, even in the ICC, where the Rules of Procedure and Evidence are firmly in the control of the States Parties, the ICC judges are permitted to adopt provisional rules where the Rules of Procedure ‘do not provide for a specific situation before the Court’ and are also permitted to adopt ‘the Regulations of the Court necessary for its routine functioning’.¹⁰²

D. Specifying an Interpretative Approach and the Functions of a MIC or MAM

Another set of control mechanisms that will be exercised collectively by all States involved in negotiating the Statute of a MIC or MAM concerns whether the Statute may specify an interpretative approach to be applied by the tribunal, and how it will specify and limit the functions of the tribunal.¹⁰³ For example, as one of the concerns about the creation of a MIC or MAM is that such a body may engage in developing a common approach across IIAs beyond what can be justified by the customary rules of treaty interpretation,¹⁰⁴ States Parties may want to remind expressly a MIC or MAM in its Statute that it should take account of relevant differences between IIAs.¹⁰⁵ More aggressively, the designers of a MIC or MAM could seek to limit the precedential weight to be given to the tribunal's decisions, eg by providing that decisions shall only be treated as persuasive in later cases in relation to IIAs containing identical treaty language.¹⁰⁶ Likewise, if States Parties are concerned to limit the increased law-making power of a MIC or MAM they may also wish to borrow from relevant aspects of the MPIA, eg specifying that the tribunal shall not address issues that are not necessary for the resolution of the dispute.¹⁰⁷

Given persistent concerns over regulatory space in investment treaty arbitration, a MIC or MAM's statute might also codify a notion of deference, perhaps drawn from existing arbitral jurisprudence, eg instructing the MIC or MAM in assessing alleged breaches of an IIA to bear in mind the ‘high level of deference that international law accords to Contracting Parties with regard to [the development of domestic policies as well as implementation of international commitments …] [and] the right to regulate’.¹⁰⁸ In designing the appellate tier of a MIC, or a stand-alone MAM, States also face, and have begun working through, a series of choices that will limit (or expand) the potential for appeals, eg which types of decisions appeals are permitted in relation to, and whether appeals are limited to errors of law or also cover certain (eg manifest) errors in fact-finding.¹⁰⁹

E. Control over Enforcement of Decisions and Remedies before a MIC or MAM

It is likely that the statute of a MIC or MAM will be drafted so that, similar to the ICSID Convention, States Parties to the statute will be obliged to recognize awards of the tribunal as binding, and to enforce pecuniary remedies in their territories as if they were a final judgment of a domestic court.¹¹⁰ This reflects the widely accepted view that for a MIC or MAM to be effective, and sufficiently attractive to investors, its decisions will need to be enforceable in domestic legal systems, similar to the existing system of investor–State arbitration.¹¹¹ Regarding the enforceability of decisions in non-Parties to a MIC or MAM, the decisions of a MIC or MAM may be enforceable under the New York Convention,¹¹² and it is also possible that non-Parties may agree to enforce decisions of the tribunal in their territories.¹¹³ In short, a MIC or MAM is likely to remain at the strong end of the enforcement spectrum, thus underscoring the need for adequate mechanisms for State input and control.

If States were particularly concerned to limit the interpretative power delegated to a MIC or MAM, they could conceivably provide in the constitutive instrument that the decisions of the tribunal only become binding after they have been adopted by the Conference of the Contracting Parties. If this approach were pursued, much would depend on the voting rules applying to adoption of decisions. For example, if adoption of tribunal decisions were, as in GATT dispute settlement, subject to consensus this would give rise to unilateral vetoes—where a single State could block the adoption of decisions—and undermine the effectiveness of a MIC or MAM. Accordingly, this would not be an advisable design choice. In contrast, if the adoption procedure were modelled on the ‘reverse consensus’ requirement in WTO dispute settlement—requiring consensus to block report adoption—it is unlikely that such a requirement would operate as a significant means of State control. Another option would be to enable some specified proportion of the Contracting Parties to block the adoption of decisions, or to determine that an interpretation shall not operate as a future precedent—ideas that have previously been proposed in relation to WTO dispute settlement.¹¹⁴

Another set of design questions concerns how the States that create a MIC or MAM may regulate the remedies that the tribunal may award. For example, following existing IIA practice,¹¹⁵ it is likely that States may limit a MIC or MAM to awarding monetary damages or restitution of property with an alternative of paying damages in lieu of restitution, and prohibit the tribunal from awarding punitive damages.¹¹⁶ More interestingly, a MIC or MAM statute might also place restrictions on how the tribunal may calculate any damages. For example, given existing concerns about this aspect of investor–State arbitration, a MIC or MAM statute might provide that the tribunal ‘shall only award monetary damages that are established on the basis of satisfactory evidence and that are not inherently speculative’ and set out the factors that the MIC or MAM must consider in assessing any damages.¹¹⁷ Overall, it would seem advisable to include such limitations in a MIC or MAM statute.

IV. BILATERAL OR UNILATERAL CONTROL MECHANISMS

This section considers those control mechanisms that will remain exercisable jointly by the parties to the particular IIA at issue or that will be exercised unilaterally by individual parties to an IIA. The analysis in this section supports the assertion in this article that despite the procedural law in a MIC or MAM having been multilateralized, certain key control mechanisms will remain exclusively in the hands of the parties to the relevant IIA.

A. Authoritative Interpretations, Treaty Amendments and Treaty Termination

The power to adopt authoritative interpretations of an IIA, including in response to prior decisions of a MIC or a MAM, will almost certainly remain solely in the hands of the parties to the particular IIA.¹¹⁸ This reflects the fact that provisions that are common in newer IIAs, which permit the treaty parties to adopt an interpretation of the agreement that binds investor–State arbitral tribunals or other dispute settlement bodies constituted under the treaty, require the agreement of all treaty parties.¹¹⁹ Similarly, subsequent agreement and subsequent practice of the parties to a treaty, pursuant to Articles 31(3)(a) and (b) of the Vienna Convention on the Law of Treaties (VCLT), require ‘a common understanding regarding the interpretation of a treaty which the [treaty] parties are aware of and accept’.¹²⁰ Likewise, the power to amend an IIA, including in response to prior decisions of a MIC or MAM, will remain exclusively in the hands of the parties to the particular IIA concerned.¹²¹ The power to terminate an IIA jointly, and neutralize the effects of any survival clause, will also remain solely in the hands of the parties to the relevant IIA.¹²²

While the power to adopt authoritative interpretations of an IIA will almost certainly remain a decision in the hands of the parties to the relevant IIA, as explained further below in Section V.D, in designing a MIC or MAM, States should create a formal process whereby all Member States of the MIC or MAM could propose, or register support for, written counter-interpretations issued in response to decisions of the tribunal, which could be applied by agreement to other IIAs besides the treaty invoked in the original proceedings.¹²³ As Mehranvar and Johnson have suggested, there are also other reforms that States could consider adopting if they want to strengthen authoritative interpretations and subsequent agreement as tools of interpretative control. These could include providing that a non-disputing treaty party's silence shall be interpreted as agreement with the host State's position or strengthening the obligation of non-disputing treaty parties ‘to engage constructively with their treaty partner(s)’ over issues of treaty interpretation.¹²⁴

B. Control over the Jurisdiction of a MIC or MAM

Another control mechanism that will operate at a more localized level is the power to decide which categories of disputes, under which IIAs (or other instruments), the tribunal is given jurisdiction over. Obviously, all States involved in negotiating the statute of a MIC or MAM will be involved in negotiating the provisions regulating the tribunal's jurisdiction. However, such provisions are likely to leave most of the detail—in terms of opting-in to jurisdiction—to be determined by individual States Parties to the MIC or MAM. For example, in the UNCITRAL secretariat Draft Statute of a Standing Mechanism, the provisions on the jurisdiction of the first-instance and appellate tribunals leave the power to grant these bodies jurisdiction up to individual Contracting Parties. Specifically, the provisions provide for jurisdiction where the disputing parties have agreed in writing to submit the dispute to the relevant tribunal and also permit a Contracting Party to the Standing Mechanism to consent to the jurisdiction of either tribunal by providing a list of treaties or foreign investment laws over which it accepts the tribunal's jurisdiction.¹²⁵ Some States have suggested that the MIC or MAM's jurisdiction should be limited to instances where all parties to an IIA have accepted the tribunal's jurisdiction over claims under the IIA, to ensure consistency of treaty interpretation.¹²⁶ Even if this were made a requirement, it would mean that the decision whether to grant the MIC or MAM jurisdiction would rest with two States (typically the parties to a bilateral IIA)—ie still only a small subset of the parties to a MIC or MAM.

V. BALANCING THE INTERESTS OF THE PARTIES TO AN IIA WITH SYSTEMIC INTERESTS

This section considers areas where there is a tension between the interests of the parties to an IIA in controlling the interpretation of their treaty, and the wider systemic interests of other Member States of the MIC or MAM that may be affected by proceedings before the tribunal. It argues that third-party participation in proceedings before a MIC or MAM—both by States that are parties to the IIA invoked but non-parties to the MIC or MAM, and by States that are parties to the MIC or MAM but not parties to the relevant IIA—is an area where the tension between procedural multilateralism and substantive bilateralism will be particularly acute. Besides third-State participation, this section also considers any right that is created to comment on draft awards of a MIC or MAM, or any forum that is created for MIC or MAM Member States to discuss, and potentially override, the tribunal's interpretations, which have been suggested as potential control mechanisms and would also raise the tension between substantive bilateralism and procedural multilateralism.

A. Participation by Non-Disputing Parties to the IIA

A strong case can be made for the right of States that are parties to the IIA at issue in proceedings before a MIC or MAM to make submissions on issues of treaty interpretation. Numerous IIAs already provide for non-disputing treaty parties to make submissions regarding issues of treaty interpretation in investor–State proceedings under the IIA,¹²⁷ and Article 5(1) of the UNCITRAL Rules on Transparency in Treaty-based Investor–State Arbitration also recognizes such a right.¹²⁸ While these existing bases for non-disputing Party participation are drafted with investor–State arbitration in mind, there is no principled reason why they should not also apply to proceedings before a MIC or MAM. Furthermore, it is likely that the MIC or MAM's statute will provide that the tribunal shall follow the UNCITRAL Transparency Rules.¹²⁹

One issue that will arise—assuming it is not made a jurisdictional requirement that all parties to an IIA are parties to the MIC or MAM—is that the non-disputing treaty party may be a party to the relevant IIA relied upon for State consent in proceedings before a MIC or MAM but not be a party to the MIC or MAM's statute. A State's non-participation in the MIC or MAM does not seem to be a good reason for denying that State the right to make a non-disputing Party submission when an IIA to which it is party is at issue before either body. While there may be a concern that non-parties to the MIC or MAM would not have contributed to the tribunal's budget, but would have some involvement in the proceedings (in making non-disputing Party submissions),¹³⁰ such concerns could be managed by orders setting the parameters for such participation, eg requiring the non-member of the MIC or MAM to pay a proportion of costs attributable to its participation.¹³¹

B. Third-Party Participation by States that are Parties to the Statute but not the IIA

The more challenging question in terms of third-party participation is whether States that are parties to the statute of the MIC or MAM, but are not parties to the IIA that is invoked as a basis of the tribunal's jurisdiction, should have any right to participate in proceedings. There are at least two reasons why such a third State may need to be afforded some right to participate in the proceedings. First, it is likely that litigation will arise over the interpretation of the statute of a MIC or MAM and the tribunal's procedural rules. Arguably all States Parties to a MIC or MAM would have a legitimate interest in how such questions are resolved and hence may need to be afforded a right to make submissions, despite not being a party to the IIA invoked as a basis of jurisdiction. Second, as investment treaties contain many broadly similar concepts and standards,¹³² and as one of the motivations for creating a MIC or MAM is to bring greater consistency of approach to the interpretation of IIAs,¹³³ it is foreseeable that positions the tribunal takes in proceedings in relation to one IIA may have significant implications for the IIAs of other State Parties to the MIC or MAM, as such a body is likely to find its own case law persuasive in later disputes.¹³⁴ Indeed, Bungenberg and Reinisch suggest that a MIC might be instructed in its statute to prefer interpretations that favour consistency and coherence in the interpretation of IIAs, a step that would obviously increase the systemic importance of its decisions.¹³⁵

A good example of an issue of systemic importance for numerous IIAs that will very likely come before a MIC or MAM is the content of the customary international minimum standard of treatment (MST) and whether that standard has evolved from the seminal Neer formulation with which it is often equated.¹³⁶ While the wording of an applicable IIA would influence precisely how the issue of the MST is relevant, the MIC or MAM's views on the content of the MST, which it could be expected to find persuasive in later cases, would have systemic relevance for all Parties to the MIC or MAM, including those not party to the IIA invoked in the initial case.¹³⁷

The question then becomes, should States that are parties to the statute of the MIC or MAM, but not parties to the relevant IIA at issue in a case, have some right to make submissions where the proceedings raise issues of systemic importance, with relevance for many other IIAs? The EU's main submission to Working Group III suggested that:

it should be considered whether and, if so, under what conditions other governments that are party to the instrument establishing the standing mechanism should be able to intervene in disputes on questions of interpretation of systemic importance under treaties to which they are not contracting parties, while ensuring at the same time that this does not compromise the ability of the parties to an agreement to retain control over its interpretation.¹³⁸

Bungenberg and Reinisch also suggest that a MIC's statute or procedural rules should provide that ‘a MIC Member who demonstrates a legal interest in a pending dispute can be admitted by the MIC as an intervening third party’.¹³⁹ The reference to a ‘legal interest’ appears inspired by Article 62 of the ICJ Statute which provides for intervention by a third State where the State has ‘an interest of a legal nature which may be affected by the decision in the case’.¹⁴⁰

The problem with a procedure for third-party participation before a MIC or MAM modelled on Article 62 of the ICJ Statute is that it would not extend to providing a sufficiently broad right of participation for a State that is a party to the MIC or MAM, but not a party to the IIA invoked in a case. The requirement in Article 62 that a State must have ‘an interest of a legal nature which may be affected by the decision in the case’ has been interpreted restrictively, and it is well established in the ICJ's jurisprudence that a general interest of a third State in the law to be applied by the Court is not sufficient.¹⁴¹

A more promising model to consider in designing a MIC or MAM may be the approach developed in WTO dispute settlement to third-party participation by WTO members besides the disputing parties. The DSU provides that any WTO member ‘having a substantial interest in a matter before a panel and having notified its interest to the DSB … shall have an opportunity to be heard by the panel and to make written submissions to the panel’.¹⁴² The requirement of a ‘substantial interest’ in a matter before a panel has not been interpreted in a manner that restricts the ability of other WTO members to intervene and, as Iwasawa notes, ‘[a] general interest in the interpretation of the WTO Agreement or a systemic interest in the dispute settlement procedures is considered sufficient’.¹⁴³ A good example is the panel proceedings in the United States – Section 301 Trade Act case, where the EU's claims largely concerned the consistency of the United States’ (US) trade laws with the WTO DSU—ie litigation over the constitutive instrument—and some 11 WTO members made submissions as third parties that largely focused on the systemic implications of the dispute for the WTO dispute settlement system.¹⁴⁴ In drawing on WTO practice regarding third-party participation it must be remembered that unlike the WTO, even under a MIC or MAM the investment treaty regime will not feature uniform treaty text on issues beyond dispute settlement.¹⁴⁵ This may mean it is appropriate to develop certain additional limitations on third-party participation by a State that is party to the MIC or MAM statute but not the IIA invoked—eg perhaps requiring such a State to show that it has at least one IIA with identical wording to the IIA at issue in the proceedings in which it wishes to make a submission.¹⁴⁶ Were such a limitation introduced, further thought would be needed regarding how it might apply when the focus of the proposed intervention was a point of customary international law, or a general principle of law, which is in turn relevant to the interpretation of IIAs, eg the question of the content and evolution of the MST mentioned above.¹⁴⁷

Ultimately, there is no easy answer to what ‘level of interest’ should be required for a third State, which is a party to the MIC's or MAM's statute but not the relevant IIA invoked, to intervene in proceedings.¹⁴⁸ Any right afforded to such States would need limits placed upon it, so as to avoid unduly expanding and burdening the proceedings before the tribunal.¹⁴⁹ Before WTO panels, third parties that notify their interest receive the submissions of the parties to the first panel meeting, and are given an opportunity to make written submissions to the panel and to be heard by the panel at a session of the first panel meeting dedicated to hearing such third-party views.¹⁵⁰ The GATT/WTO practice appears to be to disregard issues that are only raised by third parties and not the disputing parties.¹⁵¹ This may also be an appropriate limit in proceedings before a MIC or MAM, to prevent third parties which are only parties to the statute but not the specific IIA invoked from unduly expanding the proceedings. Much like existing rules for submissions by non-disputing treaty parties in investor–State arbitration, a MIC or MAM would have the power to regulate the extent of submissions of other States Parties to the MIC or MAM to ensure fairness to the disputing parties.¹⁵²

Some readers might question whether States that are parties to the MIC or MAM statute but not to the IIA invoked as a basis for jurisdiction should have any right to intervene on issues of systemic importance (eg concerning the construction of the MIC or MAM statute, or concerning standards or concepts common to numerous other IIAs). As noted above, in the ICJ and ITLOS, a general interest in the interpretation of the Statute of the Court or Tribunal and its procedural rules, or a general interest in the legal principles to be applied, is clearly not sufficient to warrant intervention,¹⁵³ even though the Court or Tribunal is very likely to follow its interpretation in later cases. In the author's view, these examples are a less appropriate model for third-State participation before a MIC or MAM than the GATT/WTO approach to third-party participation for at least two reasons. First, a MIC or MAM is likely be a tribunal with a high volume of litigation, where States are largely respondents who bear most of the litigation risk, thereby increasing the systemic interest of other Member States in how the MIC or MAM's statute is interpreted. Second, the substantial similarities in the standards and concepts found in numerous IIAs means that it is artificial to pretend that other members of a MIC or MAM will not have a significant interest in how the MIC or MAM interprets such standards and concepts, given the likely impact on later disputes under other IIAs in which the MIC or MAM can be expected to find its earlier interpretation influential.¹⁵⁴

Finally, and most challengingly, it can be asked whether States that are not parties to the MIC or MAM statute or the IIA invoked as a basis for the tribunal's jurisdiction should be afforded any right to make a third-party submission given the potential for the MIC or MAM's interpretation of relevant terms or concepts to influence later decisions of investor–State arbitral tribunals. In this scenario, one view would be there is no convincing case for providing a right to make a third-party submission because there is nothing to link such non-parties to the MIC or MAM (unlike the situation addressed above where an IIA to which they are a party comes before the MIC or MAM). A contrasting view would stress that the MIC or MAM's decisions are likely to prove influential before subsequent investor–State arbitral tribunals, potentially having some sort of superior quality, thus weighing in favour of some ability for such non-parties to make a submission.¹⁵⁵ Ultimately, it seems unlikely that the States that create a MIC or MAM would be willing to provide a basis for non-parties to the tribunal to participate in cases beyond the limited scenario where the non-party is a party to the IIA relied on as a basis for jurisdiction, given that such involvement would extend proceedings and undermine incentives to join the MIC or MAM.¹⁵⁶

C. A Right to Comment on Draft Awards

A further control mechanism where there could be a tension between the interests of the treaty parties or the disputing parties and the wider systemic interests of all Member States of a MIC or MAM concerns any right created to comment on draft awards. Such a right is provided for in a small number of newer IIAs, which, following US IIA practice, establish that, at the request of either disputing party, an investor–State arbitral tribunal shall circulate its proposed decision on liability to the disputing parties and the non-disputing treaty party or parties.¹⁵⁷ Under these provisions, the disputing parties—but not the non-disputing treaty party or parties—are given an opportunity to submit written comments to the tribunal concerning any aspect of its proposed award, which the tribunal is obliged to consider before issuing its final award.¹⁵⁸ One of the purposes of such procedures is ‘to permit both treaty Parties an opportunity to make their views known as to the impact of a proposed award on issues of public interest’¹⁵⁹ and issues of treaty interpretation, and, potentially, to coordinate and express a joint view disapproving of the tribunal's interpretation.¹⁶⁰ Coe highlights that the treaty parties may react to the circulation of a draft award containing an interpretation of an IIA they disagree with by adopting an authoritative interpretation that binds the tribunal.¹⁶¹ Several commentators have highlighted a right to comment on draft awards as a control mechanism that might be further developed in the investment treaty regime, including in a standing appellate mechanism.¹⁶² For example, Yackee suggested a stronger version of such a mechanism may require automatic circulation of draft awards to all treaty parties and permit non-disputing treaty parties to submit comments,¹⁶³ or permit the parties to an IIA to veto jointly law-making aspects of an award with which they disagree.¹⁶⁴

Were such a right to comment on draft awards created, arguably there could be a rationale for extending it to all parties to a MIC or MAM for the reasons developed above—ie, that a MIC or MAM's decisions may concern the interpretation of the tribunal's statute and procedural rules or address issues with systemic relevance for numerous other IIAs.¹⁶⁵ However, such a right to comment on draft awards sits uneasily with judicial independence and accordingly it does not seem advisable to include this control mechanism in the statute of a MIC or MAM. This is because the aim of submitting comments on draft awards would be to alter the reasoning or conclusions in draft awards after the tribunal has engaged in its deliberations. In contrast, third-party submissions made before the tribunal has engaged in its deliberations, considered above, do not raise the same concerns regarding judicial independence.¹⁶⁶

D. A Forum for Member States to Discuss Awards and Propose Counter-Interpretations

Rather than a right to comment on draft awards, an alternative mechanism to give States Parties to a MIC or MAM an opportunity to provide periodic input on the tribunal's jurisprudence may be to create a dedicated forum for this purpose, which is disconnected from the adjudicatory process in specific disputes. For example, in the WTO context, various proposals have been made to enable members to voice their views on adjudicatory decisions in a manner that is detached from report adoption in specific disputes, eg via an annual meeting of the DSB for this purpose.¹⁶⁷ Compared to a right to comment on a draft award, the advantage of a periodic forum for members to comment on the MIC or MAM's jurisprudence and operation is that it would be less clearly aimed at asking adjudicators to reverse their decision or reasoning in a pending case. However, there would still be significant tension between the influence of such a political forum and the value of judicial independence. It may also be unclear what (if anything) a MIC or MAM should do in response to comments of States Parties, for example where there are a variety of views expressed.¹⁶⁸

A related, more advisable idea would be to create a formal procedure whereby the States Parties to a MIC or MAM could propose, and indicate support for, written counter-interpretations that could be issued in response to the new standing tribunal's interpretations. In the WTO context, Fukunaga has suggested creating a formalized procedure whereby members could circulate written interpretative declarations in response to adjudicatory interpretations, essentially as a ‘counter-proposal’.¹⁶⁹ Some such statements might achieve sufficient acceptance to constitute subsequent agreement among the parties to an IIA, or alternatively have more limited weight as a supplementary means of interpretation.¹⁷⁰ Jarrett has in rough terms suggested something similar in the context of a MIC or MAM, arguing there ‘should be a multilateral forum [created] for treaty parties to annually discuss adjudicative law produced by the appellate tribunal’ and to propose changes to the tribunal's precedents, which would apply to the IIAs between all States that support the proposed change to the adjudicative precedent.¹⁷¹ What these two proposals highlight is that it is important to go beyond merely providing for a regular forum in which MIC or MAM Member States can discuss the tribunal's jurisprudence, through the creation of a procedure whereby Member States can propose, and indicate support for, counter-interpretations that effectively respond to the tribunal's interpretations. Such a procedure would amount to creating a centralized, structured process for treaty parties to do something they are already entitled to do, namely publish post-adjudication interpretative statements, including in a coordinated manner. As the plenary forum in which such counter-proposals would be considered would involve all States Parties to the MIC or MAM, it would recognize the reality, emphasized above, that the issues raised may be of systemic importance for the IIAs of numerous States, while still leaving each Member State to decide whether to apply a proposed counter-interpretation to its IIAs.¹⁷² If such a mechanism were created, it would be important to consider how written counter-interpretations, once adopted by two or more Member States, would apply to pending disputes, as permitting counter-interpretations to apply to disputes at an advanced stage would raise concerns about judicial independence and States being permitted to adopt retroactive amendments.¹⁷³

Ultimately, this proposed mechanism for circulation of, and registering support for, written counter-interpretations in response to decisions of a MIC or MAM, would create a meaningful avenue for State input and control without introducing significant risks of unilateral vetoes or abuse. This is because a written counter-interpretation would only constitute a subsequent agreement where two or more Contracting Parties agreed to apply it to the IIA(s) between them, whereas a counter-interpretation only supported by an individual treaty party would receive limited (if any) interpretative weight. Also, as noted above, limits would be placed on applying counter-interpretations retrospectively to disputes at an advanced stage. This proposal for creating a process for States to issue written counter-interpretations is consistent with States’ broader efforts within the Working Group III process to create ‘tools through which they can monitor and manage the investment treaty system to shape its evolution and emergence’, which will outlive the UNCITRAL process.¹⁷⁴

VI. CONCLUSION

This article has analysed the design of control mechanisms in a future MIC or MAM. A MIC or MAM will operate on a plurilateral basis, among States that choose to become parties to the tribunal's constitutive instrument and grant it jurisdiction over disputes under their IIAs. This article has highlighted two fundamental tensions that underlie and animate the many specific questions of institutional design. The first is the tension between independence and accountability, which is relatively familiar from other international tribunals. The second is the tension between procedural multilateralism and substantive bilateralism, which arises because a MIC or MAM will only involve multilateral agreement over dispute settlement procedures, and the relevant substantive law will remain contained in mostly bilateral investment treaties, controlled by the parties to those agreements. As this article has demonstrated, while certain key control mechanisms will continue to operate at a bilateral level, in the hands of the parties to the relevant IIA—eg authoritative interpretations, treaty amendments—others will be exercised by all parties to a MIC or MAM. The article has argued that while a MIC or MAM should include adequate mechanisms for Member State control, given the increased interpretative influence of the new centralized tribunal, it is important that such control mechanisms are designed so that they will not unduly undermine the independence of the tribunal.

A core contribution of this article has been to work though the tension between procedural multilateralism and substantive bilateralism which will arise in a more acute manner with the development of a MIC or MAM. In short, under a MIC or MAM, while the procedural law will have been multilateralized (in the tribunal's statute and procedural rules), the substantive law will remain largely bilateral. However, a MIC or MAM's decisions will be likely to have significant implications for numerous IIAs beyond the specific treaty at issue, given the similarity in content across many IIAs, the potential of the MIC or MAM to find its own interpretative approach persuasive in later cases,¹⁷⁵ and the increased weight the new centralized tribunal's decisions can be expected to carry.¹⁷⁶ This article has provided answers to consequent questions of institutional design, including how to regulate third-State participation in proceedings before a MIC or MAM—given the likely relevance of the tribunal's interpretations for numerous IIAs beyond the specific treaty at issue—and how to create a process whereby all MIC or MAM Member States could propose, or indicate support for, written counter-interpretations that respond to the tribunal's interpretations.

Besides addressing numerous questions of institutional design, this article has also provided a conceptual framework for understanding these issues that can assist policymakers considering the reform options of a MIC or MAM even if the specific approaches suggested here are not adopted. Through its analysis of control mechanisms in a MIC or MAM, this article also contributes to wider debates in international (economic) law over striking an appropriate balance between international judicial independence and Member State control. The article's analysis provides additional support for recent observations made by other commentators, who, reflecting on current ISDS reform processes, have variously noted that there are no perfect solutions in designing a MIC or MAM and all options involve trade-offs,¹⁷⁷ that the different reform options are interdependent,¹⁷⁸ and that ‘it is … proving difficult to move from a culture of bilateralism in investment arbitration to one of multilateralism’.¹⁷⁹