CETA’s Investment Court System and the Autonomy of EU Law
On 26 June 2018, the European Court of Justice (ECJ) held its hearing in Opinion 1/17, which concerns the compatibility of CETA’s Investment Court System (ICS) with EU law, in a Full Court sitting. The Opinion was requested by Belgium; the Council, the Commission and twelve Member States intervened in the oral proceedings. All intervening parties except for Slovenia argued that CETA is fully compatible with the EU law. Somewhat surprisingly, the European Parliament chose not to submit observations, although it previously backed the conclusion of CETA.
Opinion 1/17 is the latest episode in a series of Opinions and judgments concerning the interrelationship between international courts or tribunals and EU law. Most of them did (initially) not end well. In Opinion 1/91, Opinion 1/09, Opinion 2/13 and, most recently, Achmea, the Court found that the EU membership in or accession to an international dispute-settlement system breached, above all, the principle of autonomy of the EU legal order.
This short blog post seeks to give some insights from the hearing in Opinion 1/17. Due to the depth of the substantive issues covered and the length of hearing, it only addresses the first of four questions posed to the ECJ by Belgium, which relates to the principle of autonomy. It is fair to say though that autonomy-related questions and arguments featured the hearing most prominently and may therefore be particularly contentious.
The arguments by the parties relating to the autonomy criterion
In its first question, Belgium seeks to ascertain whether CETA’s ICS complies with the principle of autonomy of EU law, including in particular the CJEU’s exclusive jurisdiction to interpret EU law.
Article 28.31 paragraphs 1 and 2 CETA include four main ‘autonomy safeguards’. First, the provision stipulates that the CETA Tribunal shall only apply the provisions of CETA, read in conjuncture with the Vienna Convention on the Law of Treaties and other rules of international law applicable between the Parties. Second, it includes an absolute prohibition of the CETA Tribunal to determine the domestic legality of a measure. Third, in determining whether a measure violates CETA, the Tribunal may only consider domestic law ‘as a matter of fact’, following the prevailing interpretation given by domestic courts or authorities. Fourth, the meaning given to domestic law by the CETA tribunal ‘shall not be binding upon the courts or the authorities of that Party’.
In essence, the answer to the first question posed by Belgium therefore depends on whether the safeguards stipulated in Article 28.31 CETA are sufficient for guaranteeing the uniform and consistent interpretation of EU law and the principle of autonomy of the Union legal order.
The submissions by the Council and the Commission
Opinion 1/17 marks one of the few CJEU decisions where Council and Commission stand on the same side. All intervening Member States except Slovenia supported the submissions of the Council and the Commission.
The Council emphasised, first, that the CETA tribunal must not interpret or apply EU law. Article 8.31(1) CETA would ensure that the CETA tribunal’s jurisdiction is limited to the interpretation and application of CETA provisions alone. Unlike the ECHR agreement (Opinion 2/13) or the EEA I agreement (Opinion 1/91), CETA would not touch upon EU primary or secondary law but sets its own standards. The one and only task of the CETA tribunal was therefore to determine the compatibility of EU measures with the standards enshrined in CETA, considering EU law only as a matter of fact. In its oral submissions, the Commission added that the mere possibility that an international court may have to ascertain, as a matter of fact, the meaning of a domestic EU law provision with a view to ruling on its compatibility with CETA, could not per se be incompatible with the principle of autonomy, provided such an assessment would not have the effect of binding the EU and its institutions in the exercise of their internal powers. Otherwise, it would be impossible for the EU to remain in or adhere to virtually any international dispute settlement mechanism.
Secondly, the Council and Commission, in their respective submissions, underlined that CETA does not prevent the use of the preliminary reference mechanism specified in Article 267 TFEU or the rights of Union courts to hear cases that concern substantive investor rights enshrined in EU law. Instead, CETA leaves a choice for Canadian investors to use either the ‘domestic court route’ or the ICS. The two routes ought to be considered mutually compatible as the exclusion of direct effect under Article 30.6 CETA would provide for a divide, as it were, separating CETA standards from the EU legal order. Investors may only use the ICS to claim alleged breaches of CETA which they cannot invoke in front of domestic Union courts (due to the lack of direct effect). Rights enshrined in Union law, by contrast, may only be enforced by bringing a case to the Union courts – and ultimately the CJEU – as the CETA tribunal must not rule on the legality of domestic law (Article 8.31(2) CETA).
Thirdly, the Council stressed that CETA would leave the primacy of EU law in tact as the EU would only be bound by the outcome of an ICS decision (i.e. financial compensation) but not by the meaning given to domestic law by the CETA tribunal (i.e. the reasoning). In contrast to the European Court of Human Rights, for example, the CETA tribunal cannot set aside EU law.
Fourthly, the Council recalled that contrary to the situation in Opinion 2/13, Art.8.21 CETA read together with Regulation 912/2014 prevents CETA from pre-defining any issue of competence delineation with regard to determining a respondent on the side of the EU or financial responsibilities arising from an award.
Lastly, the Council and the Commission both argued that the Achmea judgment, in which the Court held that the investor-state dispute settlement (ISDS) mechanism enshrined in intra-EU bilateral investment treaties (BITs) is in violation of the principle of autonomy of EU law, could not be extrapolated to the extra-EU investment treaty context. CETA’s ICS would differ significantly from the ISDS mechanism prevalent in the Dutch-Slovakian BIT which was relevant to the Achmea case: Article 8 of that BIT confers upon the ISDS tribunal the competence ‘to take into account … the law in force of the contracting parties’ as such and not merely as a matter of fact. Moreover, the principle of mutual trust – which underpinned the Court’s reasoning in Achmea – would not apply between the EU and a third state, such as Canada.
The submissions by Slovenia
Whilst Belgium, in its oral observations, merely questioned the compatibility of CETA’s ICS with EU law, Slovenia flat-out rejected it. In its written observations, Greece shared the concerns of Slovenia but decided not to make an oral submission.
In essence, Slovenia argued that Art.8.31 CETA does not provide sufficient safeguards to guarantee that the autonomy of EU law is left intact. In particular, Slovenia submitted that in assessing EU law as a matter of fact the ICS tribunal would in some situations necessarily have to interpret the substance of EU law. Moreover, Slovenia underlined that CETA lacks a definitive guarantee to ensure the functioning of the preliminary reference mechanism as domestic proceedings must be stalled as soon as a case is brought under the ICS (Article 8.22(1)(f)-(g) CETA). In addition, the exclusion of direct effect would grant the CETA tribunal in certain areas exclusive jurisdiction.
The questions by the Court relating to the autonomy criterion
In the second part of the hearing, eight Judges and Advocate General Bot posed a variety of questions to the Council and the Commission. It is beyond the scope of this short blog entry to recall the entire dialogue. Instead, the following will merge some questions (and responses) relating principally to the autonomy of the EU legal order criterion in two thematic groupings.
Scope of jurisdiction: relevant vs. applicable law
Various questions posed by the Court and the Advocate General related to the scope of the CETA tribunal’s jurisdiction. Advocate General Bot commenced his questions to the Commission with a request for terminological clarification: What is the precise difference between ‘applicable’ and ‘relevant law’ as stipulated in Article 8.28(2) CETA? The Commission replied that the term ‘applicable law’ specified in paragraph 2(a) of Article 28 CETA refers to the interpretation of CETA whereas the reference to ‘relevant domestic law’ made in paragraph 2(b) of the same provision concerns errors of facts in the interpretation of domestic law.
In addition to and, in some instances, building on this legal terminological clarification, several questions by the Court and the Advocate General related to the reach of the CETA tribunal’s powers to consider the ‘relevant domestic law’. Re-occurring case examples by the Judges centred on the EU legislatures’ right to regulate (Article 8.9(1)-(2) CETA) and the CETA tribunal’s jurisdiction to evaluate and limit the EU’s powers to impose environmental standards on Canadian investors: What is the applicable law for Canadian investors in the EU? Do EU regulations on the environment fully apply to Canadian investments within the territory of the Union?
The Council answered the latter question in the affirmative – presuming that there is a legitimate policy objective on the side of the EU legislature. The CETA tribunal would hence assess the EU environmental measure applicable to Canadian investors ‘as a matter of fact’. This response, however, raised further questions as to the scope of the CETA tribunal’s factual assessment of EU law, especially as the Tribunal may have the capacity to strike a different balance between environmental and economic values than the Union legislature and courts. To paraphrase a couple of questions posed by ECJ President Koen Lenaerts to the Council and the Commission: Could the CETA tribunal simply disregard EU law as interpreted by the CJEU as a matter of fact, finding it unreasonable and not really squaring with the EU’s conferred powers in the field of the environment? Would the CETA tribunal be bound by the CJEU’s interpretation of the ‘relevant law’ in its assessment of the ‘applicable law’?
In its reply, the Council emphasised that the ‘right to regulate’ was drafted with the intent to limit the CETA tribunal’s jurisdiction, granting the domestic legislature the broadest possible discretion. Fiercely underlining that CETA ensures that the Tribunal must not interpret EU law and evaluate its domestic legality, the Council and the Commission stressed on several occasions that the Tribunal’s powers are strictly limited to determining whether an EU measure is in violation of the standards set in Sections C or D of Chapter 8 CETA. In its assessment of the ‘relevant law’, the Tribunal would have to take account of CJEU decisions.
Choice of courts
A second recurring issue was the choice granted by CETA to Canadian investors to bring a claim either in front of domestic Union courts or the ICS. Would such a choice not lead to ‘forum shopping’? Can a decision by a Union court, and ultimately the CJEU, be subsequently ‘overturned’ by the CETA Tribunal? Could a ‘choice of courts’ lead to double compensation? Does the exclusion of direct effect grant the CETA tribunal de facto monopoly powers?
In response to the Judges’ questions, the Council repeatedly reiterated that although the CETA leaves the Canadian investor a ‘choice of courts’, the legal bases for each court’s decision and the remedy it might impose differ. If an investor were to claim that an EU measured violates rights guaranteed by CETA, she could – due to CETA’s lack of direct effect – only bring a case in front of the CETA tribunal with the intention of obtaining financial remuneration. If, on the other hand, the investor would claim that a national or EU measure breached rights guaranteed to her by EU law, she would choose the ‘domestic court route’ seeking an annulment of the measure in question. CETA would in any event prohibit double compensation or punitive damages. In view of the complementarity of the two different ‘court routes’ one could neither speak of ‘forum shopping’ nor would the ICS have a monopoly position as investors may always choose to exhaust local remedies. In the course of the hearing, the Commission furthermore accentuated that the Court of Justice would always retain ex-post facto control irrespective of the finalité of the award levied by the CETA tribunal. As the meaning given to domestic law by the CETA tribunal is not binding on Union courts, the ECJ could still rule not to execute the award should it arrive at the conclusion that it is incompatible with EU law. The Member States would in such a scenario be bound by the primacy of EU law. The consequence would be retaliation by Canada.
The bigger picture: What is at stake?
Quite remarkably, the hearing was not purely legal. Recurring policy-related comments and questions underlined the near impossibility to separate the legal arguments set forth in Opinion 1/17 from the political ramifications of the Opinion’s outcome.
In the course of the hearing, several Member States, the Council and the Commission urged the Court to consider the need to cultivate the development and strengthening of a rule-based international order, especially in current times. The autonomy of the EU legal order principle should not be interpreted so narrowly as to prevent the EU from remaining in or adhering to any international dispute resolution mechanisms. Smashing the ICS – and with it the endeavour to establish a Multilateral Investment Court – would be a step backwards, potentially leading to a re-nationalisation of external investment policy at Member State level and a regress to the ISDS system.
Indeed, the Opinion might have a decisive impact on the architecture of EU trade and investment agreements. In the aftermath of Opinion 2/15 on the FTA with Singapore, the Council has decided to follow the Commission’s proposal to split the exclusive ‘EU-only’ trade part of EU FTAs from the ‘mixed’ investment protection part on a case-by-case basis. The separation of trade and investment disciplines on the side of the EU counters the general trend at international level to conclude comprehensive trade and investment agreements. The provisional application of the ‘EU-only’ part of comprehensive agreements such as CETA has same practical effect. Opinion 1/17 could lead to a ‘definitive splitting’: If CETA’s ICS system were to be held incompatible with the EU legal order, it might be impossible for the EU to conclude investment protection agreements at all, leading to a separation between external EU trade and internal EU investment policy. Beyond the realm of investment, Opinion 1/17 will surely set new legal parameters for the EU’s membership in or accession to international courts and tribunals. If the ECJ were to hold CETA’s ICS incompatible with EU law, the EU institutions can only hope that the decision includes exhaustive criteria offering clear guidance on the policy changes required for rendering the ICS compatible with EU law after all. This approach has recently been followed in Opinion 1/15 (para 232). Otherwise, the EU’s future accession to international (arbitral) courts or tribunals might fade into the far distance.