Many arbitration lawyers’ initial reaction to the CJEU’s Achmea judgment resembles the first three of the famous “five stages of grief” (denial, anger, bargaining, depression and acceptance). Some deny Achmea’s relevance under international law, others angrily dismiss it as unreasoned and politically motivated, while many attempts to “bargain” a way out for intra-EU arbitrations under the ECT and/or ICSID given their multilateral and extra-EU character.

In my view, Achmea is entirely consistent with both EU and international law, and applies equally to intra-EU ISDS under BITs, the ECT and ICSID. It may, therefore, be wisest to keep the stage of depression brief and accept the outcome before moving on to greener pastures.

The Autonomy and Primacy of EU Law

To restate several truisms, the TEU and TFEU ( “EU Treaties”) are inter-State agreements that establish an autonomous system of regional international law. The fundamental purpose – indeed, the raison d’être – of this system is to create a common market without internal frontiers. To this end, the CJEU has developed the doctrines of “primacy” and “effectiveness”, under which any domestic legal rule – even one that enjoys constitutional status – that is incompatible with EU law must be immediately disapplied, without awaiting formal revocation [C-6/64, Costa v. ENEL [1964] ECR 585; C-106/77, Simmenthal [1978] ECR 629]. The doctrines of primacy and effectiveness are long-standing and considered by all as an integral part of the EU Treaties.

Inter se agreements of EU Member States are not exempt from primacy and effectiveness, and have always been treated just like domestic law [E.g. C-235/87, Matteucci [1988] ECR 5589, paras 19-22]. This is logical because EU Member States cannot be permitted to deviate from common market rules by treaty any more than they can by domestic legislation. It is also consistent with the international treaty law, which expressly permits States to

(a) disapply provisions of an earlier treaty to the extent they are incompatible with those or a later treaty [VCLT Article 30(2)-(3)],

(b) enter into subsequent agreements concerning the interpretation and application of their previous treaties [VCLT Article 31(3)(a)], and

(c) subject their international agreements to other “rules of international law applicable in the relations between the parties” [VCLT Article 31(3)(c)].

In other words, there is nothing legally wrong with EU Member States agreeing, in the EU Treaties, that those Treaties will automatically prevail over all other inconsistent international agreements between those same Member States. There is no sense in denying the international character of such an agreement or its relevance under international law.

No “Outsourcing”

Like virtually all cross-border economic activity inside the EU, intra-EU investments take place squarely within the scope of the “four freedoms” of the EU’s internal market. Many areas in which investment disputes tend to arise (e.g., energy, environment, state subsidies) are heavily regulated by internal market directives, regulations and/or state aid law. Furthermore, Member State conduct is already subject to binding EU standards of protection of property rights, due process and legal certainty (including legitimate expectations), which are all relevant to the legitimacy of a particular government measure. Much as tribunals may seek to avoid them, questions of EU law therefore inevitably arise – indeed, centrally so – in many intra-EU arbitrations.1)See e.g., Electrabel v. Hungary Decision on Jurisdiction and Liability, paras 6.21-6.22, 6.70, Micula v Romania Award, para 691-707, 741-742, 792-793; Wirtgen v Czech Republic Award, paras 297-299, 337-342, 350; Postova Banka v Greece Award, paras 192-193, 207-208.

Given the fundamental importance of EU law primacy and autonomy, the CJEU always firmly opposed Member States’ “outsourcing” to non-EU fora disputes even potentially involving their EU law obligations. This was the case with the EEA Court in Opinion 1/91, inter-State arbitration in the Mox Plant case, the Patent Court in Opinion 1/09, and the European Court of Human Rights in Opinion 2/13. The Achmea judgment refers to this line of case law and is fully consistent with it.

Advocate-General Wathelet proposed treating ISDS tribunals as Member State courts, enabling them to refer EU law issues to the CJEU. That would have been an elegant solution indeed if it were not for one obstacle: convincing ISDS tribunals themselves. How can one be sure that a tribunal seated outside the EU and composed of non-EU lawyers would even recognise an EU law issue, let alone suspend proceedings for a CJEU reference?

It is, of course, correct that commercial arbitration is permitted under EU law, and commercial awards’ compliance with EU law can be verified by EU courts at the post-award stage [See e.g. C-126/97 Eco Swiss [1999] ECR I-3055]. However, parties to a commercial contract are not under a legal duty to ensure the primacy and full effectiveness of EU law in a certain territory. By contrast, Member States conferring jurisdiction on an ISDS tribunal do have that duty, and the remote possibility of post-award control by EU courts where the tribunal may not even be seated in the EU is far from sufficient to secure compliance.

There is thus nothing shocking or “political” about the Achmea judgment. The dismay that ISDS lawyers now express towards the Achmea judgment mirrors similar dismay expressed four years ago by human rights lawyers towards Opinion 2/13. However, like that Opinion, the Achmea judgment is only a restatement of the fundamental constitutional principles that have been at the heart of CJEU’s jurisprudence for many decades.2)See e.g., D. Halberstam, ‘It’s the Autonomy, Stupid!’ A Modest Defense of Opinion 2/13 on EU Accession to the ECHR, and the Way Forward, 16 German Law Journal 105 (2015)

Multilateral Agreements

There have been suggestions since the Achmea judgment that it may not apply to ICSID and/or ECT arbitration because those treaties include third countries.

However, the CJEU has always been clear that primacy also covers any intra-EU application of multilateral conventions [See e.g. C-301/08 Bogiatzi [2009] I-10185, paras 3-4, 19.]. This is consistent with TFEU Article 351, which regulates the relationship between EU law and international treaties involving third countries, as well as with VCLT Article 41(1)(b), which permits States to vary multilateral treaties as between some parties, to the extent this does not affect the rights of others.

The language of the Achmea judgment very clearly extends to all provisions “such as” Article 8 of the Czechoslovak-Dutch BIT. That includes the ISDS provisions of the ECT, which is, therefore, extinguished, as between EU Member States, by operation of primacy. Similarly, primacy will “turn off” the ICSID Convention to the extent it applies to intra-EU ISDS awards. Given that such application of primacy does not affect the interests of any non-EU States or their investors, neither the ECT nor ICSID can provide a safe ground on which intra-EU ISDS can continue. The CJEU is quite uncompromising and there is no place for bargaining here.

A Question of Trust

It would be unthinkable for an investor from Berlin (London, Paris) that is denied a planning permission in Munich (Manchester, Marseille) under German (British, French) administrative law to oppose the refusal outside the framework of that law, before a foreign arbitrator unfamiliar with that law, on the vague grounds that the refusal was “unfair” or “inequitable”. The relationship between investors operating within a single national market and the state that regulates that market is normally governed by the administrative law of that state based on the principle of equal justice for all. After all, Texas and California cannot agree on special privileges for each other’s investors independently and outside of US federal law.

Why should things be different within the European single market? The answer often given to justify this anomaly is that the courts of certain recent joiners to the EU are somehow inadequate compared to those of “old” EU Member States. Notably, no evidence of this appears in existing intra-EU awards, virtually none of which concerned any kind of judicial impropriety. However, the assumption that there is a risk of such impropriety does shed light on another ground on which the CJEU decapitated intra-EU ISDS: a breach of the EU principle of “mutual trust”.

After comprehensively reforming their legal system to meet the EU’s stringent accession criteria, the courts of “new” Member States are deemed to meet the standards of justice that all Europeans are entitled to expect. There are no “second-class” Member States in the EU and the principle of mutual trust implies a presumption that a Dutch investor in Slovakia can expect substantially the same quality of justice as it can in its home Member State.

If, as many complain, this is not the case, in reality, that would undoubtedly be a failure of the EU. However, it would be one that must be rectified at Union level, and in a manner that complies with Union law. ISDS lawyers can and should play a role in arriving at a solution, but the first step must surely be to understand and accept the reasons and consequences of the Achmea judgment, not deny, dismiss or try to bargain a way out of them.

Although the author is involved as counsel in investment disputes, the views expressed in this blog entry are the author’s alone and are not to be attributed to any person the author represents or has represented in the past.


________________________

To make sure you do not miss out on regular updates from the Kluwer Arbitration Blog, please subscribe here. To submit a proposal for a blog post, please consult our Editorial Guidelines.


Kluwer Arbitration

References

References
1 See e.g., Electrabel v. Hungary Decision on Jurisdiction and Liability, paras 6.21-6.22, 6.70, Micula v Romania Award, para 691-707, 741-742, 792-793; Wirtgen v Czech Republic Award, paras 297-299, 337-342, 350; Postova Banka v Greece Award, paras 192-193, 207-208.
2 See e.g., D. Halberstam, ‘It’s the Autonomy, Stupid!’ A Modest Defense of Opinion 2/13 on EU Accession to the ECHR, and the Way Forward, 16 German Law Journal 105 (2015)
This page as PDF

6 comments

  1. Good analysis. But on this analysis why only intra-EU BITS? EU law is as much at issue whenever an EU MS is a respondent regardless of the nationality of the claimant, no?

    1. @Lorand Bartels Existing extra-EU agreements are protected from EU law by Article 351 TFEU – see, e.g. Case C-264/09 European Commission v. Slovak Republic [2011] ECR I-8065, where the ECJ expressly exempted the Swiss investor Atel from the Second Electricity Directive because sanctioning ATEL under the Directive would be an expropriation under the Switzerland-Slovakia BIT and the ECT.

      As regards NEW investment treaties (whether signed by the EU or by individual Member States), after Achmea, it would presumably be an infringement of EU law to include investment arbitration provisions in there without sufficient safeguards for the autonomy of EU law. Let’s await the court’s opinion on CETA.

      1. A nuance on my above comment is of course that Article 351 TFEU only covers pre-accession extra-EU agreements and would not, on its face, apply to extra-EU agreements concluded by a State while it is a member of the EU (which is the case, for example, for many Dutch, German, French, Spanish and Italian BITs with third countries). This category of BITs would require a separate thorough analysis, but it is clear for example that they do not call into question the principle of mutual trust, which does not operate vis-a-vis third countries; the applicability of Article 344 is also not evident (it may be read as requiring an agreement between at least two EU Member States). Furthermore, Opinion 2/15 (on the EU-Singapore FTA) confirmed that Member States have external competence with regard to ISDS (i.e. they are free to agree ISDS mechanisms with third countries, within the bounds of EU law). My sense is that even if the BITs falling into this category are incompatible with EU law, they would not automatically be turned off by primacy, because primacy of EU law only applies as between EU Member States, not EU Member States and third countries. The EU Member States would, however, have the obligation to amend or terminate them. I think this issue too should be clarified by the forthcoming CETA opinion.

  2. Sir,

    I agree with your analysis. It occurs to me that the Member States that are contracting parties to an EU-internal are obliged to comply with the ruling in Achmea as of 6 March 2018. Therefore the proceedings under which an investor from one of those Member States may, in the event of a dispute concerning investments in another Member State, bring proceedings against the latter Member State before an arbitral tribunal whose jurisdiction that Member State has undertaken to accept, have become inapplicable.

    In addition, the Member States that are parties to such agreements are required to denounce the arbitration clauses contained in the BITs as soon as possible. Furthermore, the maintenance of this dispute resolution mechanism will engage the responsibility of the Member States concerned. Finally, as the guardian of the treaties, the European Commission should initiate infringement actions in accordance with Article 258 TFEU against any Member States that are not willing to denounce these ISDS clauses.

    Given that there are 181 intra-EU BITs in force, the Achmea judgment is likely to have a significant impact on these treaties. Almost all of these agreements have been concluded between old and new Member States.

    What is more, although the Achmea case concerned an investment treaty concluded between two Member States, specifically the Netherlands and Slovakia, the implications of the judgment in question reach far beyond internal EU agreements. In fact, the dispute resolution mechanisms provided for under BITs concluded between Member States and third countries will also prove to violate Article 267 and 344 TFEU where the arbitral jurisdiction is liable to concern either the application or the interpretation of EU law.

    Nicolas de Sadeleer,
    Jean Monnet chair
    Professor of EU law, St Louis Univ.

  3. A lot of comment overlooks the fact that the CJEU’s judgment was only on the compatibility of the arbitration mechanism of the BIT with EU law, and not of the substantive investor protections under the BIT. This is important for at least three reasons.

    – Firstly, because most BITs give the investor a choice between domestic court jurisdiction and arbitration. An investor can still therefore invoke the substantive protections of an intra-EU BIT before the domestic courts of the member state concerned.

    – Secondly, because it admits a “quick fix”, through amending the treaty (or just domestic law) to ensure that disputes over EU law can be referred to a domestic court and thence to the CJEU, but with all non-EU law points remaining within the sole jurisdiction of the tribunal.

    – Thirdly, because it is not obvious that this limited degree of incompatibility is sufficient under the Vienna Convention to deprive the arbitration mechanism of effect in international law. Article 30 of the Vienna Convention only appliues to successive treaties “relating to the same subject matter”; query if the TFEU has the “same subject matter” as a BIT. If that is right, then an investor has merely to constitute a tribunal with a seat outside the EU; while any resulting award might not be enforceable within the EU, a state that refused to comply would be at risk of having its foreign assets seized.

    The Achmea judgment also overlooks that the nature of the arbitral mechanism under a BIT is not an agreement between two states for the arbitration of investment disputes, but rather a standing offer made by each party to the treaty to investors – i.e. private individuals and companies – to arbitrate disputes. In this respect, it is conceptually very difficult to distinguish an arbitration between a private entity and a state under a BIT from a commercial arbitration to which a state is party, and the CJEU’s rationale for doing so (namely, that a commercial arbitration results from the free decision of the parties to contract out of the domestic courts) is utterly unconvincing: what is the difference, from the perspect of enforcing EU law, between a state signing a contract with an investor agreeing to arbitrate a dispute arising from the investment, and offering to do so via a treaty?

    1. I agree that substantive protections of the BITs are left in place. In many EU Member States which subscribe to the monist regime, international treaties are given direct effect and can be (and often are) invoked in State courts. However, EU courts will give primacy to EU law, and if they fail to do so, be sanctioned by the CJEU. That is not the case with ISDS tribunals, which is precisely why the CJEU found them to be incompatible with EU law.

      On the Vienna convention – I do not believe it enjoys some kind of constitutional or ius cogens status. EU Member States have agreed among themselves that EU law prevails among all other inconsistent treaties between them, regardless of the subject matter. Even if that departs from the Vienna Convention (which is debatable), it is the States’ good right to do so.

      On the contractual arbitration involving the State point – the of course State remains free to contract with investors subject to its internal legislation (which itself must be compatible with EU law) and State aid rules. If the contract violates EU law, an arbitral award enforcing it may be challenged or denied enforcement under the ECO Swiss doctrine. By contrast, a BIT places the State’s sovereign conduct (administrative, legislative and judicial) under a supervisory regime that is parallel to and independent of EU law and may, at times, be directly contrary to it. To me, the difference is quite obvious.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.