This paper has been prepared in response to the Commission's consultation regarding a proposed multilateral investment court system. It has been prepared in this format rather than through the Commission's questionnaire in order that some more general comments can be made relating to the Commission's proposals and the premise on which a number of the questions have been based. However, where this paper addresses specific questions raised in the consultation, these have been clearly flagged in the footnotes. The response is divided into four sections. The first sets out our understanding of the Commission's proposals for either a Multilateral Investment Court System (M-ICS) or a Multilateral Appeals Tribunal (MAT). The second considers the scope of the Commission's efforts to establish a new framework (of either the M-ICS or MAT) and the potential political and practical implications. The third section considers the benefits and impact of an M-ICS of MAT within the context of the differing proposals. The fourth section considers the characteristics which may be considered advantageous for either system and the potential differences which might be anticipated, depending on the choice of an M-ICS or MAT. Operating from over 26 offices across Asia Pacific, EMEA and North America, Herbert Smith Freehills is one of the world’s leading professional services firms. We provide many of the world’s most important organisations with access to market-leading dispute resolution, projects and transactional legal advice, combined with expertise in a number of global industry sectors, including energy, natural resources, infrastructure, technology and financial services. We have a marketleading reputation for commercial arbitration, investment treaty arbitration and public international law practices, with specialist practitioners globally. We represent the Kingdom of Spain in defending two claims brought under the Energy Charter Treaty. Other state clients have included Tunisia, Malaysia and the US Department of Justice. We also represent a number of corporate clients in pursuing investment treaty claims against states. Our practitioners regularly sit as arbitrators and are appointed to roles at the major arbitral institutions. As a firm we are actively involved in policy-making for both states and corporate counsel. 1. MULTILATERAL REFORM OF THE INVESTMENT DISPUTE SETTLEMENT SYSTEM: TWO POTENTIAL OPTIONS Current EU policy is to include in each EU trade and investment agreement an institutionalised procedural framework for resolving investment-related disputes by establishing an Investment Court System (ICS) for each agreement. This approach has been adopted following previous consultations regarding Investor State Dispute Settlement (ISDS) and we understand that the ICS is considered by the Commission to address a number of the shortcomings identified with the preexisting ISDS system. This approach has since been adopted in the EU-Vietnam and CETA agreements. The Commission has indicated its intention to continue seeking inclusion of some form of ICS in its future bilateral trade and investment agreements with third party states. However, this consultation focuses on whether, in parallel, the EU should expand its efforts and pursue the establishment of some form of new multilateral system for resolving investment disputes. From the consultation it would appear that the Commission is considering two potential options: 1. A new M-ICS formed of a first instance and appellate level "court", which would replace certain existing forms of investor-state dispute settlement. 2. Retaining existing forms of investor-state dispute settlement, but introducing a new appellate level tribunal, the MAT, which would sit above the existing forms of dispute settlement and seek to introduce consistency through the appellate body, rather than through the first instance procedure. If the M-ICS is the preferred system, the Commission seeks to consult on whether: 11/41406155_5 2 a) The M-ICS should replace the individual ICSs established under the EU's existing agreements (such as the EU-Vietnam FTA and CETA). b) Member States should also seek to amend the dispute resolution provisions of their Bilateral Investment Treaties (BITs) with third states to refer disputes to the M-ICS. c) The Commission should also lobby for non-EU states to adopt the M-ICS for resolving disputes under their investment treaty arrangements. The Commission then seeks to consider what characteristics are considered important in relation to the M-ICS. As an alternative, the Commission proposes the adoption of a MAT. If the MAT is to be the method chosen, the Commission seeks to consult on whether: a) The MAT should hear appeals from the individual ICSs established under the EU's existing agreements (such as the EU-Vietnam FTA and CETA). b) Member States should also amend the dispute resolution provisions of their BITs with third states to introduce the MAT as a new appellate level. c) The Commission should lobby for non-EU states to adopt the MAT as a new appellate level in the ISDS process. The Commission then seeks to consult on what characteristics are important in relation to the MAT. 2. THE SCOPE OF ANY NEW MULTILATERAL INVESTOR-STATE DISPUTE RESOLUTION SYSTEM 2.1 Achieving Global reach The Commission recognises in the consultation that there are limits to what can be achieved by the EU and its Member States on a global basis. However, the consultation questions whether the EU can and should seek to achieve an over-arching investor-state dispute resolution system which would be able to hear disputes between investors and states, whether contracting with the EU or not, on a bilateral and multilateral basis. The EU has previously consulted on its own approach to ISDS and reached a conclusion that the introduction of an ICS is the best approach for its own future agreements. However, it is not apparent that those previous consultations sought to consider the broader global picture of investor-state dispute resolution and whether, on a global basis, that system is effective. Clearly, there will be a number of significant political and practical obstacles to seeking change on a global scale, even assuming that sufficient other states share the EU's concerns regarding the existing system. There will also be varied opinions about how best such concerns should be addressed. For example, ICSID has already been established as a global body to address investment disputes between states and investors. It is a multilateral dispute mechanism, adopted in many of the world's investment agreements and the Convention has been signed by 160 states. Certain concerns have of course been expressed about the current ICSID system, for example in relation to the annulment process, timescales, and the concept of precedent. While a small number of states have recently sought to leave ICSID, it still remains a system with wide international acceptance and buy-in, with an established enforcement process. While the enforcement leverage the World Bank has been questioned by critics, there is no doubt that the World Bank's involvement has force. It is not yet clear whether the systems contemplated by the Commission would be designed so as to be sufficiently coordinated and consequential that they would offer the same kind of leverage in the event of non-compliance with an award. ICSID is currently consulting on revisions to its arbitration rules. We consider that in the first instance all interested parties should engage in this process with the aim of addressing existing concerns with the current system, including the EU and the Commission. This seems to us to be the correct next step, rather than looking for the adoption of a wholly new system and infrastructure, particularly at this time given the political challenges facing many new and existing international trade and investment treaties worldwide. 11/41406155_5 3 2.2 Extension to the bilateral arrangements of Member States The consultation also requests views regarding whether it is possible, or desirable, to extend the M-ICS or MAT to Member States' bilateral investment agreements. In particular, the Commission questions whether consistency between the national and European level is important.1 We do not consider that such consistency is important, and certainly not at this early stage. It is self-evident that Member States' BITs with third countries, many of which were concluded some time ago, will not include reference to an ICS system. The existence of pre-existing dispute resolution arrangements should not therefore be undermining the EU's negotiating position. In any event, we understand that existing bilateral arrangements with EU members will be terminated following signature of an EU-level agreement FTA. Furthermore, as a practical matter we doubt that it will be possible to achieve re-negotiation of existing BITs on any significant scale in the near future. As the Commission itself acknowledges, it "is not conceivable that such a high number of investment treaties could be renegotiated to allow [sic] to make changes to the ISDS provisions".2 The consultation suggests that a convention similar to that of the Mauritius Convention might be a way forward. A multilateral convention could of course achieve such a result in principle, but again this is likely to be difficult to negotiate given different political priorities and viewpoints. The Mauritius Convention itself took many years to negotiate, and affects the procedure to be adopted between a state and an investor once an arbitration has been commenced, not the method of dispute resolution itself. It has also only been ratified by two states to date.3 In this regard, this difficulty in amending Member States' BITs would seem to apply equally to both the M-ICS and the MAT. The MAT does not appear to us to be any more straightforward to adopt than a new M-ICS, since both would likely require a re-negotiation of existing BITs and the acceptance of a new dispute resolution body (notwithstanding the implication given by question 45 of the consultation). The MAT ought therefore to be considered on its own merit, in our view, and not because it may be a more easily achieved system of reform. 2.3 Extending a M-ICS or an MAT to the EU's existing and future investment agreements The question of whether the EU is able to negotiate the agreement of past and future signatories to Investment Agreements with the EU to either an M-ICS or an MAT is a political one. It may well depend on whether the proposed multilateral arrangement offers the third party state the same benefits as the existing individual ICS contained in their current arrangement. In terms of the M-ICS, this may well depend on the format to be offered. For example, under the CETA, the "court" that is formed will always contain a Canadian decision-maker. In a multilateral arrangement, this benefit may be lost.4 Similarly, the ICS system contained within the CETA considers payment of a stipend to court members, but does not obviously envisage the establishment of an infrastructure or facilities for that court. As discussed below in section 3.1, while the Commission has not clarified what the M-ICS would look like, it is likely that there would be a greater cost ascribed to that arrangement for a third party state that is signatory to only one of the EU's investment arrangements, than a more ad hoc ICS system. While subject to a number of similar concerns, an MAT may be more achievable in the context of the EU's current and future Investment Agreements which contain an ICS. These already envisage the establishment of an appellate court, the details of which remain to be finalised. It may therefore be easier for the EU to seek consistency at this appellate level across all its Investment Agreements than to focus on harmonising those agreements by creating an entirely new dispute resolution system. 1 Questions 28 and 29. 2 Introduction to Question 28. 3 See, as an example, that the Mauritius Convention has, to date, only been ratified by two states, Canada and Mauritius. http://www.uncitral.org/uncitral/uncitral_texts/arbitration/2014Transparency_Convention_status.html 4 See Questions 47 and 53. 11/41406155_5 4 3. BENEFITS AND IMPACT 3.1 Benefits and impact of a multilateral system 3.1.1 A cost benefit? The possible costs benefits of either system will depend on the Commission's intentions in terms of scope (as discussed above) and on the type of body being considered. For example, it is difficult to answer whether it will be more cost effective to maintain an M-ICS than a separate ICS for each of the EU's investment agreements, as it will depend heavily on the type of "court" envisaged. The larger the "court", the more likely it is to need headquarters, infrastructure and a secretariat.5 The cost of maintaining such an establishment to deal with, what is, in reality, a limited number of disputes, may be difficult to justify. Paying a limited number of "adjudicators" a retainer under separate agreements, allowing them to be called to deliberate on an ad hoc basis without the need for a formal establishment to house them may allow the parties to avoid the costs of a standing establishment and result in less cost and bureaucracy. Broader economic impacts are also difficult to ascertain with certainty. For example, introducing an M-ICS over all of the EU's agreements with a system of binding precedent and an appellate structure, may, over time increase predictability and reduce cost. However, since the number of cases brought under the same treaties on the same legal or factual points may be small, there will likely be a relatively long period of time where the costs involved for both states and investors are higher, as parties seek to appeal more decisions until a consistent body of jurisprudence is developed. Meanwhile, introducing an MAT (depending on its scope on law and fact) may increase cost over the existing system. Only very limited grounds for challenge are permitted under the New York Convention for non-ICSID awards and annulment challenges under ICSID may only be brought on specific grounds. The Commission should also consider the additional cost that would rest with states in maintaining two different multilateral systems of dispute resolution under both ICSID and either an M-ICS or MAT (depending on the scope of the system), 6 and whether the benefits of the M-ICS or MAT are sufficient to warrant that additional cost. There is also the risk of the "political" cost of an MAT that is relatively inactive. Given that the grounds for challenge under the New York Convention are so limited (and relatively speaking, an MAT may be less active than ICSID annulment committees), there is the risk the MAT creates a subsidiary set of decisions which are marginalised- in turn undermining the significance of the MAT and its jurisprudential presence. Moreover, it is unclear on what basis referral to the MAT would occur, since any prevailing investor will go straight to a national court of enforcement. Of course, consistent with the standard commercial arbitration model, the existing ad hoc arbitration system preferred by other states worldwide for their investment disputes would appear to be the most cost-efficient, since the cost is borne directly by the users. 3.1.2 Contributing to the global economic climate? The Commission questions whether the introduction of an M-ICS or MAT may contribute in a positive way to the Global Investment Climate. 7 The availability of non-national dispute resolution and protections under a treaty may help to foster investment, but the introduction of a major new, untested system is unlikely to affect the global investment climate positively, at least initially. Depending on their format and scope, either system may help to strengthen the legitimacy of ISDS, but this remains to be seen. 3.1.3 Consistency and predictability? A further benefit of the M-ICS and MAT systems identified by the Commission relates to the increased predictability and consistency in investment dispute resolution. 8 Central to this must be the introduction of a binding system of precedent under either system. In this respect there are 5 As noted in Question 54. 6 See 4.8 below regarding protection of developing countries. 7 Question 56. 8 Throughout, but including Question 43 and 44. 11/41406155_5 5 questions in relation to both the M-ICS and the MAT: how will the system of precedent work? Will only the decisions at appellate level in the M-ICS be considered as binding (and then only upon the first instance tribunals)? Would decisions at appellate level in the M-ICS and in the MAT bind only in relation to the specific treaty being considered in that relevant decision, with potential influential value (but short of amounting to binding precedent) in similar language in different treaties? Centralisation and the use of binding precedent may help in bringing clarity and certainty to language in particular treaties. However, it is worth noting that, again, depending on the potential scope of the system established, this may take many years to achieve given the limited number of cases being brought under the same treaties. Furthermore, the consistency benefit depends again on scope. A system which all EU Member States sign up to (but not their BIT counterparts) and all states which have signed trade agreements with the EU, will just create a second "centralized" system, alongside ICSID, while retaining other forms of ad hoc or institutional arbitration. Separately, consideration must be given to the role of EU law in any such EU institution. For example, there remain unresolved debates over where EU law sits in relation to customary international law. The application of international law to Investor-State disputes will presumptively be governed by the lex specialis of any BIT/MIT/IIA, however, the interplay with EU law will be an important foundational element to ensuring consistency and predictability. 4. DESIGN, COMPOSITION AND FEATURES OF A SINGLE MULTILATERAL INVESTMENT COURT OR A MULTILATERAL APPEAL TRIBUNAL 4.1 Possible features of an M-ICS or MAT The features which should be seen in an M-ICS or MAT will depend on the Commission's decision on the intended scope of such an arrangement.9 The Commission lists a number of possibilities. These include: • Permanent dispute resolution structure • Appeal instance to correct errors of law and manifest errors of fact • Full-time adjudicators • Fixed remuneration for adjudicators • High qualification criteria for selecting adjudicators • Random allocation of cases • Transparency / full documentation disclosure requirements • High ethics standards • Safeguards for independence (e.g. random allocation, tenure, etc.) For a truly global M-ICS or MAT or one which encompasses the bilateral arrangements of Member States, a permanent structure with full-time adjudicators would seem a sensible approach. Yet for the more limited application of an M-ICS, or particularly, MAT, to the EU's agreements, the cost of maintaining a permanent structure with full-time adjudicators may be prohibitive. For the reasons discussed above, the random allocation of cases may also lessen the likelihood of third party states agreeing to give up their individual existing ICS arrangements. Many of the other qualities listed above may well be necessary based on a value decision as to what should or should not be expected of adjudicators in an M-ICS or MAT, and the basis of their appointment and remuneration. This element is discussed in section 4.3 below. 9 Questions 30 and 31. 11/41406155_5 6 4.2 Appointment, challenge and removal of Adjudicators The Commission asks whether it would be important that each country referring disputes to the MICS or MAT be able to appoint an adjudicator to the panel.10 In the CETA text, the EU and Canada each appoint adjudicators to the ICS established under it. Again, depending on the scope of either arrangement, the agreement referred to it and the position on whether or not nationals of a country are excluded from hearing that dispute, it would be important to clarify whether or not Member States would be entitled to appoint their own adjudicators. As noted above in 4.1, a third state which already has the ability to name a number of adjudicators to the panel and guarantee that a national of that state will hear the dispute, may be reluctant to give up this prerogative. The Commission also asks whether the number of adjudicators should be tailored to the likely number of cases and not linked to the number of countries signatory to the agreement. 11 Again, the response to this question depends on what the structure of either system would be. With a global M-ICS with an established secretariat and infrastructure and standing body of adjudicators, each country may be willing to pay to ensure that it is represented on the panel of adjudicators. If an MAT for the EU's investment agreements is established, then it would be more likely to need to be based on the number of potential cases. Regarding the selection of arbitrators, the ICS in the CETA assumes the appointment of a panel by the State parties. This differs substantially from the current selection process for arbitrators under other Investment Agreements. If the aim of the Commission is to improve confidence in the ISDS process, it will be important to ensure that investors feel that the adjudicators are truly independent and impartial of their state appointees. A strong ethics requirement will be critical, along with a robust system to challenge and remove an arbitrator for breaches of independence and impartiality. 4.3 Qualifications, remuneration and independence of adjudicators in both systems We agree with the Commission's proposed criteria for eligibility to sit as an adjudicator under either system, particularly expertise in public international law and previous experience in international investment law12. In terms of remuneration,13 what is appropriate will depend also upon the restrictions placed on the adjudicators once in office. Given the high level of qualifications required, it is unrealistic to suggest that the M-ICS or MAT system will be able to attract the right calibre of individuals to sit as adjudicators and require that they not to work as experts or counsel or publish academic or legal papers unless they are paid a consistent monthly income. This is particularly the case in the early stages while cases are still relatively rare, as being paid per case would be unlikely to generate sufficient income. 14 It therefore rests with the Commission to determine whether safeguarding independence through these restrictions warrants the cost of retainer. This may, in turn, rest on the scope of either the M-ICS or the MAT.15 An MAT established to hear appeals from the EU's Investment Agreements will have far fewer cases than a M-ICS which has jurisdiction over all of the EU's investment Agreements and the bilateral arrangements of Member States. From a purely practical perspective, the latter may warrant a salary structure and more rigorous restrictions on its adjudicators than the former. It should also be noted that, if separate ICSs are established under each of the EU's investment agreements and each adjudicator is on a retainer linked only to that specific agreement, multiple ICS operating in parallel next to each other will raise the question of how many potential individuals can be found who would fulfil the necessary qualification and impartiality requirements. Finally, the level of activity of the institution chosen will have an impact on the candidates that may be put forward. For example, an inactive MAT might not attract (or be able to attract) adjudicators who, practically speaking, might be sitting as arbitrator in the cases that come to appeal. This runs 10 Question 46. 11 Question 47. 12 Question 49. 13 Question 51. 14 Question 52. 15 Question 51. 11/41406155_5 7 the risk that the quality (and qualifications) of the candidates selected for the MAT is lower than the first instance arbitrators. This may affect confidence in the system. 4.4 Appeal on fact and law The Consultation asks whether an appeal to correct errors of law and manifest errors of fact is a desirable feature of either the M-ICS or MAT. This is already a feature of the CETA and it is assumed, therefore, that the EU has already determined that an appeal on the basis of a manifest error of fact, as well as law, is desirable in its investment agreements. Again, therefore, the grounds for appeal under the M-ICS or the MAT depend in most part on the scope of either arrangement. Any typical appellate system requires an established body of law which can be relied upon at first instance. Given the lack of stare decisis under public international law and the sometimes disparate jurisprudence offered by ad hoc investment arbitral tribunals, it may be challenging to establish an agreed reference point from which it can be assessed that there has been a manifest error. While some view an appellate system as a form of protection for states, this ultimately assumes that the State parties are the losing parties. This is not borne out by the jurisprudence to date.16 Offering an appeal may be viewed by many states as giving investors another "bite at the cherry" once they have lost. Allowing for an appeal on the basis of an error of fact as well as law, may simply be viewed as exacerbating this risk. In the event that the M-ICS or MAT is extended only to the EU's existing and future Investment Agreements, the ability to appeal on the basis of an error of fact will be common to the existing arrangements and should be uncontroversial. For Member States' BITs and those of third states this will be a substantial, and potentially unwelcome, change. If an appeal is to be allowed on such a basis (in whatever form of system) it will be critical that the appellate system is managed robustly, and "manifest error" is interpreted clearly and consistently to mitigate the risk of investors or states using the system frivolously for unmeritorious challenges. 4.5 The need for enforcement17 The EU has retained a system of ISDS in its Investment Agreements, even if it has been changed from the existing form of investment arbitration. It is assumed that the decision to retain ISDS has been made on the belief that direct enforcement of protections by investors against states outside domestic courts is advantageous for the investment climate. On this basis, the recognition and enforcement of any resulting decision will be of utmost importance for either the M-ICS or MAT system. Offering investors a form of dispute resolution without any "teeth" will render the entire process nugatory. It will also be important to demonstrate the EU's commitment to and respect for the system.18 The current ISDS options ensure enforcement, either under the ICSID or New York Conventions. It is assumed that an MAT, regardless of its scope, would deliver decisions in the form of Awards given that the first instance decision would have been rendered by an arbitral tribunal. However, given the Commission's efforts to move away from nomenclature associated with arbitration, the question arises whether any M-ICS would need a new convention or provision for the recognition and enforcement of the decisions of the M-ICS. Given the limited uptake amongst states to both the Mauritius Convention and Hague Conventions on Choice of Court, it seems unlikely that such an enforcement convention would achieve wide international buy-in. 16 The SCC's recent report on Investment Arbitration confirmed that 58% of cases brought before the SCC were resolved in favour of the Respondent state http://sccinstitute.com/media/178174/investor-statedisputes-at-scc-13022017-003.pdf . UNCTAD's analysis of the outcome of such cases also indicates that the position is more nuanced http://investmentpolicyhub.unctad.org/ISDS 17 Questions 40 and 41. 18 See, for example, the EU's response to the Micula decision. 11/41406155_5 8 It would therefore seem appropriate and most practical for the Commission to seek to bring any proposed system within the enforcement regimes of the New York Convention or ICSID which already provide a clear system of enforcement. 4.6 Allocation of operational costs19 The Commission asks at Question 55 whether any of the "operational costs" of either the M-ICS or the MAT should be paid by user fees (either the Investor or State). The alternative option is that the entire system, including the salary of any arbitrators would be paid by the State parties. This is a substantial step away from both the current ICSID, ad hoc or institutional systems which are available for ISDS. While ICSID itself is supported by the state signatories, both the states and Investors must pay for the cost of the arbitrators hearing their case. In other forms of ISDS, the investor and state must pay for the tribunal, the fees of the relevant institution and for the hearing with the costs potentially then re-allocated depending on outcome. While ensuring there are no fees to pay other than the costs of legal representation demonstrates a very "investor-friendly" approach, the Commission has not indicated its rationale for this change. State parties invited to adopt either the M-ICS or MAT may feel it increases the risk of an unmeritorious claim if the Investor's cost burden is limited to this extent. 4.7 Protections for SMEs The Commission questions whether protection should be given in any M-ICS or MAT system for SMEs.20 Size of enterprise in the context of investor-state dispute resolution would appear a somewhat blunt tool. It does not follow that the amount being claimed by the SME will necessarily be smaller than that claimed by a larger investor: if the SME's whole investment is in the state's jurisdiction then it could claim a substantial sum. Offering a different or faster process based on the size of the enterprise is also unlikely to be welcomed by either side. An SME is unlikely to want to adopt a "lesser" process if the reduced robustness of that process increases the chance of an appeal, and the further delay and cost that would entail. Similarly, states are unlikely to view being sued by an SME as a lesser risk, or that an SME's claim warrants expedition any more than that of a larger investor. Rather than offering a reduced or shortened process based on the size of the investor, it would appear more appropriate to introduce a more flexible process based on value or complexity of the claim. However, it may be possible to limit the use of certain procedural tools such as security for costs against SMEs or the ability of states to challenge the use of third party funding against SMEs. We would suggest that having a flexible hearing location is rather dependent on what the EU envisages as the structure of any M-ICS or MAT and whether it will have its own specific premises. In any event, if a flexible hearing location is offered for SMEs, that same flexibility should apply to any proceedings. HERBERT SMITH FREEHILLS LLP 19 Question 55. 20 Question 35.