Don’t Lead with Your Chin! If Member States continue with the ratification of CETA, they violate European Union law
By Christina Eckes
After last week’s Achmea ruling of the Court of Justice (CJEU) Member States can no longer legally go ahead with ratifying CETA – the mixed Free Trade Agreement that the EU and its Member States agreed with Canada. Achmea casts serious doubts on the legality of CETA’s investment chapter, which allows investors from one Party to submit to an arbitral tribunal a claim that the other Party has breached an obligation under CETA. By simply going ahead with the ratification, they violate the principle of loyalty under European Union law.
On 6 March, the CJEU declared in its Achmea ruling that the investor-state-dispute-settlement (ISDS) mechanism in the bilateral investment treaty between the Netherlands and Slovakia (NL-SK-BIT) as incompatible with EU law. A request by Belgium is pending before the CJEU asking for clarification on the legality of the new Investor Court System in CETA (Opinion 1/17). Achmea is a clear indication that the CJEU in Opinion 1/17 is likely to find also the Investor Court System in CETA problematic for the autonomy of EU law.
No general obligation exists for Member States to halt national ratification of mixed agreements when their compatibility with EU is questioned before the CJEU. Yet, CETA is different. The clear indication of incompatibility in Achmea imposes an obligation on national Parliaments to halt the CETA ratification process and wait for Opinion 1/17.
Member States are obliged to protect the EU’s unity in international representation.
EU loyalty requires that Member States amongst others ‘facilitate the achievement of the Union’s tasks and refrain from any measure which could jeopardise the attainment of the Union’s objectives.’ In external relations, they are obliged not to undermine the EU’s external actions and ensure unity in international representation.[1] Obligations flowing from EU loyalty also stretch to and are arguably all the more necessary where the power to conclude an international agreement lies (solely) with the Member States.[2] Furthermore, EU loyalty covers not just the present state of EU law but also ‘the foreseeable future development of EU law’ and should hence be interpreted as requiring certain actions or omissions in the present in order to avoid a potential future conflict between international legal obligations and EU law.[3] It is reflected not only in the general provision on sincere cooperation, but also in several specific Treaty provisions, such as Member States’ duty to re-negotiate or even denounce prior international agreements. Under specific and very limited circumstances, this duty can even require taking measures to re-negotiate an agreement in order to avoid potential future incompatibles.
The very procedure that Belgium used to address its questions about CETA to the CJEU is meant to allow the Court to review the compatibility of envisaged EU international agreements with EU law before they enter into force in order to avoid undesirable external consequences, such as international responsibility. CETA is a so-called mixed agreement, i.e. an agreement concluded by the EU and its Member States together. Mixed agreements require ratification by all Member States and conclusion by the EU. Requesting the CJEU to give an opinion does not halt the process of conclusion by the EU per se.[4] Hence, if all Member States ratify and the EU concludes CETA, the Court would no longer be able to give an opinion on its compatibility with EU law because CETA would no longer be an ‘envisaged agreement’, but already in force. In practice, Wallonia – if it does not change its position with the new government – and Slovenia have declared that they will wait for the Court’s Opinion. However, if Member States continue to ratify CETA despite the clear warning sign in Achmea and the CJEU then finds the Investor Court System to threaten the autonomy of EU law, this is an open display of disregard for the pending opinion request, as well as a display of a lack of unity in external representation. Achmea case is a timely and open indication of serious risk that the CJEU considers the Investor Court System in CETA as incompatible with EU law. An indication that Member States cannot ignore.
Can Achmea really be read as an indication that the Investor Court System in CETA is also problematic for the autonomy of the EU legal order?
Two arguments are presented to distinguish the Investment Court System in CETA from the ISDS in Achmea: First, CETA contains a new modernized form of ISDS possibility, an Investment Court System, which meets higher rule of law standards, in particular by improving the internal coherence of investment law and the independence of arbitrators, as well as by reducing the private autonomy bias in the system. This is true as such, but the reasons for which the CJEU declared the ISDS in Achmea incompatible with EU law also apply to CETA. The CJEU introduced its ruling with 8 paragraphs of principled considerations on the autonomy of the EU legal order before turning to the questions on the compatibility of the ISDS mechanism. The autonomy of the EU law, referred to by the CJEU as justified ‘by the essential characteristics of the EU and its law’ (Achmea, para 33), is what makes EU law a legal order. Only the autonomous character of EU law, i.e. self-referential nature, not depending for its validity and interpretation on national and international law, vests it with a constitutional character and the ability to ensure the effectiveness and uniform application of EU law.
The Court then found the ISDS in Achmea to go against that autonomy. This is in line with its settled case law that EU participation in international dispute settlement mechanisms, while it remains possible as a matter of principle, may threaten the very nature and existence of the EU as an autonomous legal order. Most well-known of this line of case law is the Court’s rejection of the draft agreement on EU accession to the European Convention on Human Rights in 2014. The autonomy of the EU legal order can only be preserved by a judicial system that is capable to ensure the consistency and interpretation of EU law (Achmea, para 35). The whole system is undermined if disputes could be removed from the EU judicial system by taking them to arbitral tribunals, which do not form part of the EU judicial system and consequently cannot ask preliminary questions to the CJEU (paras 50-52). In particular the arbitration mechanism specifically interprets EU law, as the arbitration tribunal in Achmea did when it dealt with Slovakia’s objection that Article 8 of the NL-SK BIT is incompatible with EU Law (see pp 68-72). These fundamental concerns of the CJEU seem to apply in the same way to the new Investment Court System under CETA.
Second, the Court has ruled in Achmea on a bilateral investment agreement between two EU Member States and CETA is a Free Trade Agreement between the EU and all 28 Member States as one party and a third country as the other. If anything, this makes the situation more problematic. The CJEU rejected in Achmea that an arbitral tribunal established under an international agreement between two Member States could be considered part of the judicial system of the EU (Achmea, para 45). The same conclusion applies even more obviously to an arbitral tribunal established in a multilateral agreement involving a third country.
Achmea made apparent the tangible probability that the Court finds fault with CETA in Opinion 1/17. The Court’s reasoning in Achmea confirms that the threat to the autonomy is not averted by the CETA provision that aligns the interpretation of EU law by the arbitral tribunal to the ‘prevailing interpretation’ of the CJEU and excludes that the CJEU is bound by the tribunal’s interpretations.[5]
In the light of the foreseeable risk that CJEU declares the CETA investment chapter to be capable of undermining the autonomy of the EU legal order, Member States are required by the principle of EU loyalty to halt ratification in order to demonstrate a uniform position as one Party, together with the EU and the other Member States, on the international plane in general and vis-à-vis Canada in particular.
[1] E.g.: Case C-266/03, Commission v Luxemburg [2005] EU:C:2005:341, para 60; Case C-433/03, Commission v Germany [2005] EU:C:2005:462, para 66.
[2] Opinion 1/13 [2014] EU:C:2014:2292, View of AG Jääskinen, para 99.
[3] Ibid, para 100.
[4] See Opinion 3/94, Bananas agreement, EU:C:1995:436, para 23.
[5] Article 8.31 CETA; nearly identical: Section 3, Article 16(2) of the EU-Vietnam Free Trade Agreement.
12 comments
Very interesting post. However, I respectfully disagree. In a dispute based on an intra-EU BIT, EU law constitutes law applicable to the dispute (as such subject to the interpretation of the arbitral tribunal) within the meaning of Article 31(3)(c) of the VCLT. Hence the threat to the autonomy of the EU legal order. The same reasoning cannot be applied to a dispute based on an agreement concluded by the EU (and its MS) with a third country, such as CETA. In this case, EU law becomes domestic law of one of the parties to the agreements (and to the dispute), not binding on the third country in question and not directly applicable and interpretable by the ICS.
Thanks for your thoughts Luca. This is I think not how the ECJ sees it (see paras. 39-42 Achmea) and that’s the relevant benchmark. The key point is that the ISDS tribunals, including the ICS, remove disputes from the jurisdiction of the courts of the Member States and hence go against article 19 TEU. Don’t forget that CETA will become part of EU law, so that the principles of Opinion 2/13 fully apply. If you create a system that rivals with the powers of domestic courts, instead of a straightforward state to state dispute settlement system with a country outside the EU you are bound to affect the autonomy of EU law. Moreover, I would add, CETA introduces the novelty of allowing tribunals to ‘consider’ EU law as ‘a matter of fact’ and to follow only prevailing interpretations of EU law. Breaking up EU law as a matter of fact and as a matter of law might not be something the ECJ is willing to engage in, especially if this distinction is determined by ICS tribunals and not by the ECJ.
Very interesting post Christina. I think you must be right. My understanding of Achmea is that the Court considers it contrary to EU law for a MS to enter into an agreement which may cause it to be bound by decisions of a tribunal if
(i) the tribunal may take decisions affecting EU law in that state – which is almost inevitable for any trade or investment tribunal, and
(ii) it cannot refer questions to the Court, and
(iii) there is no possibility to substantively review the decisions in the MS courts.
This would be just as applicable if a MS had a BIT with a third country which involved arbitration.
And, as you say, I think it gets CETA: although this is concluded by the EU, that doesn’t seem to be the central factor in the Court’s reasoning. It is worried about how ISDS-type tribunals affect the working of EU law. Those concerns transpose perfectly to the CETA ‘court’. If the EU wants to include investment tribunals in its treaties it therefore looks as if it will have to have either references to the Court from them, or the possibility of appeals to MS courts. Which makes them a bit pointless.
RIP ISDS.
Dear Christina,
Thank you for your post. I expressed my thoughts on this judgment on another forum. I would make two comments:
1. While I understand, and I agree, that the Achmea judgment might signal the Court’s stance on the ICS in CETA as well, I would still frame this statement as a possibility not an almost given fact. The CJEU in Achmea mentions several times that tribunals “such as” the one in the NL-SK BIT are incompatible with EU law. I think “such as” is an important qualification: on the one hand, it ensure that the judgment is applicable to other similar intra-EU BIT ISDS mechanisms, on the other hand it does – in my opinion – also restrict the Achmea judgment to ISDS mechanism that are “such as” the ones in the NL-SK BIT. What does this entail though? Does this mean that it covers ISDS mechanisms in extra-EU BITs as well? Or does it mean that it applies to those ISDS mechanisms – both intra and extra-EU – in which the tribunals apply EU law? In Achmea, Article 8 of the BIT clearly said that the tribunal also applies the law of the contracting parties and the international agreements between them, thus EU law. The CETA mechanism, on the other hand, talks about the application of domestic law as fact and not as applicable law to the dispute. I am of the opinion that in essence this type of “application as fact” could still affect the uniformity of EU law, but the CJEU might decide otherwise.
2. I am not entirely convinced by the loyal cooperation argument. For footnotes 2 and 3 you cite AG Jaaskinnen. Nonetheless, I don’t remember the CJEU picking up on these two points (correct me if I am wrong). So, it is debatable whether those two obligations mentioned by the AG are a component of the loyal cooperation obligation or not. Furthermore, I don’t see why the duty to denounce an agreement is relevant to this case? Article 351 TFEU only refers to this obligation in case of prior agreements (CETA is not) and it is not applicable in intra-EU situations (such as the SK-NL BIT). Furthermore, as mentioned earlier, it is not yet sure whether the CETA ISDS mechanism is incompatible with EU law. Yes, Achmea might be a sign that it is not, but this is not certain. So, why should an MS do something if it is not yet certain about why it would need to do it? Also, this is a mixed agreement, endorsed by the EU and ratified by the EP. If the MS ratify it, what they are doing in essence is complying with the policy objectives of the Commission, the Council and the EP, instead of hindering the wishes of 3 EU institutions by not ratifying the agreement.
Dear Szilárd,
Thank you for your insightful comments.
1) The CJEU has not yet ruled on a similar situation. This is why the references in the post are to situations that impose far-reaching duties on Member States under international relations, such as denouncing an agreement that they have already ratified (Art 351 TFEU situation). I do not claim that Article 351 TFEU is directly applicable.
2) I do not want the Member States to do anything. I only want them to refrain from doing something, i.e. ratifying CETA. This would be a public and legal expression of support for an agreement that is most likely against EU law.
3) Most parts of CETA are provisionally applied anyway. So, it mainly concerns really the controversial investment chapter.
3) Indeed, a mixed agreement will not enter into force as long as not all Member States and the EU have ratified it. Yet, ratification after Achmea (this is the core of my argument) is different from the support of the EU institutions for CETA before Achmea. Achmea was a strong indication that the ICS in CETA is contrary to EU law and Member States should simply sit on their hands for a while to give the CJEU a chance to give its opinion.
Dear Christina,
Many thanks for your post. I have two questions for you:
1) What do you make of paragraphs 57 and 58 of the judgement? In your interpretation, is there not a distinction between Member States not submitting disputes outside the judicial system and the EU itself resolving its disputes with third countries by means of international adjudication? Why would the EU itself be able to decide exceptions to the principle of legal autonomy, which all countries that accept the jurisdiction of international tribunals do?
Such a restrictive interpretation would appear to mean that the EU itself would either not be able to enter into enforceable international agreements of any kind, or only recognise amicable or coercive means of dispute settlement.
– Adjudicatory dispute settlement clauses in existing agreements would be incompatible and the EU would be prevented from agreeing ad hoc to adjudicate existing disputes. If the EU could not subject itself to the WTO Dispute Settlement Understanding, the WTO dispute settlement system’s compatibility with the Treaties would also be in doubt.
– Denying itself the possibility of agreeing to settle disputes by arbitration would leave the EU with the two remaining ‘means of dispute settlement’ known to international law: amicable means (negotiation, mediation) and coercive means (unilateral economic/political forms of retaliation). That would seem an unnecessary self-imposed constraint on any international player. Certainly since other States, not uniquely the EU, also have to accept that their laws are applied or interpreted by international tribunals, such as by the WTO Appellate Body.
2) Why would the Achmea test not also put Member States’ submission to ECtHR jurisdiction (pursuant to Art. 34 ECHR) into doubt? Looking at the three prongs of the test – i) ECtHR may make decisions interpreting or applying EU law in that State, ii) ECtHR cannot make references for preliminary rulings under Art. 267 TFEU, iii) no possibility to substantively review ECtHR decisions by Member State courts – it seems that either the Achmea test must know some limits or else there may even broader implications.
To me the judgement contains indications that the Achmea ruling may not be intended to have such broad application.
Dear Jens,
The relevant distinction is between state-to-state dispute settlement mechanisms and dispute settlement mechanisms involving individuals.
The Treaties in principle permit international agreements providing for state-to-state dispute settlement between the EU and third countries (such as the WTO’s dispute settlement body). Also the question has not been brought to the CJEU at the time and the court can only rule on ‘envisaged agreements’ (Article 218(11) TFEU), so in practice the Court cannot speak to this point. The Court has time and again confirmed that in principle the EU can participate in international dispute settlement, as long as it does not undermine the autonomy of the EU legal order.
State-to-state dispute settlement is a very different cup of tea. Proceedings brought by individuals involving questions of EU law are subject to the preliminary ruling procedure in Article 267 TFEU. This procedure has been called the ‘keystone’ of the EU’s judicial system by the CJEU for a reason. Article 267 TFEU is a formal, institutionalised judicial dialogue procedure that ensures the uniform interpretation and enforcement of EU law via national courts in all 28 Member States. Article 267 TFEU was also central to the CJEU’s reasoning when it found that the Treaties constituted ‘a new legal order’ that gives individuals, not just the Member States, rights and obligations, whose uniform interpretation the CJEU oversees.
The CJEU has made clear in no uncertain terms that it has the exclusive power to give definitive interpretations of EU law and therefore ensure the uniform interpretation of EU law across Europe (See ECHR Opinion 2/13, paras. 244-248). However, a fundamental purpose of ICS in CETA is to enable investors, i.e. individuals, to challenge not only EU acts and decisions based on these acts, but also national acts which might involve EU law in some way, an ICS tribunal would necessarily remove disputes that involve EU law from the EU judiciary, consisting of national courts and the EU Courts, connected by Article 267 TFEU. In the worst case, ICS would have to interpret and give meaning to EU law. Similarly, to the context of human rights law (ECHR Opinion 2/13), ICS would therefore encroach on the powers of the EU Courts, either by removing disputes from the Article 267 TFEU mechanism or by ruling on questions of EU law. Furthermore, ICS in CETA does not require the exhaustion of domestic remedies or prior involvement of the CJEU, which would soften the risk of divergent interpretation, as well as respect the powers of the courts of the Member States to hear claims by individuals involving questions of EU law.
Many thanks for your explanations. But does the Art. 267 argument really resolve the issue? A State-to-State arbitral tribunal could not be a ‘court or tribunal’ of a MS under Art. 267 for the reason you mention (267 refs for prel rulings do not emanate from State-to-State disputes). But that just leads to the conclusion that a State-to-State arbitral tribunal does not fall outside the scope of Art. 344 as being a ‘method of settlement other than those provided for [in the Treaties]’. In other words, we are still left with the problem of squaring a MS’ submission to State-to-State arbitration with Art. 344. As for the EU itself, even if ISDS encroaches on the principle of autonomy of the legal order, why would the EU be incapable of making exceptions to that principle, just the United States and other federal states (or any states for that matter) necessarily do when accepting the jurisdiction of an international tribunal. Would not the CETA ‘full monty’ ratification procedure suffice or would it require a change to the Treaties? Maybe we have to wait for opinion 1/17 to know the answer to that last question. Thanks again.