The Missing Link: Direct Effect, CETA/TTIP and Investor-State Dispute Settlement

 

 

By Daniel Thym, Chair of Public, European and International Law, University of Konstanz

International treaties have rarely received more attention than the proposed free trade deals between the EU and the US and Canada. This entails that many law students and practitioners are confronted with a theme that does not feature prominently in legal education. In debates with students, I realise that preconceptions about the functioning of domestic legal systems are regularly projected upon the international sphere. One example is a demand that companies should challenge state action before domestic courts instead of dispute settlement bodies under the planned EU/US agreement (TTIP) and the proposed EU/Canada trade agreement (CETA). These claims often assume that national courts hold the competence to enforce corresponding rules. For lawyers working on domestic issues it seems self-evident that courts can apply the law.

Against this background, this blogpost focuses on a provision in the Draft CETA with Canada (Article 14.14: see the text below), whose relevance has not been acknowledged so far, including by the stimulating contributions to the Verfassungsblog Symposium on Investment Protection. On page 470 of the roughly 1600 pages of the consolidated CETA Draft Agreement, which the Commission regards as a template for free trade negotiations with the United States, we come across a final provision of seemingly minor relevance on ‘private rights’, which rejects the applicability of the agreement en passant. This reaffirms that the implications of the free trade deals would be less dramatic than some suggest.

Background: CETA and TTIP as International EU Agreements

In order to understand the relevance of Article 14.14 on private rights, we should apprehend that most segments of the free trade agreements would be binding upon Member States as an integral part of EU law. Axel Flessner may try to argue that the arrangements would constitute an ultra-vires-act (thereby initiating more tweets than any other contribution to the said symposium), but the plain Treaty text demonstrates that the legal picture is fairly evident. Article 207 TFEU declares that the EU’s Common Commercial Policy (CCP) allows for the conclusion of trade agreements on goods and services and embraces, among others, ‘foreign direct investment’, while Article 3 TFEU maintains that the conclusion of agreements in this area shall be an exclusive Union competence.

These provisions were a deliberate policy choice after decades of wrangles about the precise scope of the CCP. The Treaty of Lisbon was meant to replace earlier and ambiguous formulations with an overarching competence for the European Union. In its Lisbon Judgment, the German Federal Constitutional Court recognised the pertinence of this change: ‘With the exclusive competence as set out above, the Union acquires the sole power of disposition over international trade agreements which may result in an essential reorganisation of the internal order of the Member States.’ Judges in Karlsruhe gave the green light to the changes nonetheless, albeit with a minor caveat that ‘much argues in favour of assuming that the term “foreign direct investment” only encompasses investment which serves to obtain a controlling interest in an enterprise’ and excluded so-called portfolio investments (ie, non-controlling interests in companies) as a result.

This reference to the limits of the CCP is relevant, since it indicates, in general terms, that there remain uncertainties about the precise delimitation of competences for corollary aspects of international trade. For that reason, most national governments maintain that CETA and TTIP should be concluded as so-called ‘mixed agreements’, with the EU and all 28 EU Member States as signatories. If that view prevailed, national parliaments would have to give their consent as well. However, this would not modify the internal allocation of powers; the EU institutions are in the driving seat in the vast field of Union competence – both during the negotiations and with regard to legal effects. It is established case law that the legal effects of mixed agreements follow the rules of Union law for all matters that are covered by the Common Commercial Policy.

Domestic Application of Agreements concluded by the EU

The predominance of Union law in legal practice can obscure our awareness of the specificities of the international legal system. Law students across Europe learn in their undergraduate courses about the direct and supreme effect of supranational rules, but are not always familiar with public international law. Domestic courts have to apply Union law in pretty much the same was as national law, but this assumption cannot be extended to public international law without modification, including in situations in which the EU concludes international agreements with third states.

It is true that the ECJ maintains that international agreements can have direct effect as an integral part of the Union legal order – and an example demonstrates that this can have critical implications: for example, judges in Luxembourg decided in July that Member States cannot automatically require the spouses of Turkish nationals to acquire basic language skills of the host country. Legally, this conclusion rests upon an interpretation of the so-called standstill provision for the self-employed in the Additional Protocol of 1970 to the Association Agreement between Turkey and today’s European Union. In the terminology of international trade law, the case concerned a so-called non-tariff barrier to the freedom of establishment. It is these non-tariff barriers that take centre stage in both CETA and TTIP negotiations (neither of the latter treaties will affect migration, though).

If the underlying legal arguments about direct effect and court jurisdiction extended to free trade deals with Canada and the United States, the implications could be dramatic indeed. Both the ECJ and domestic courts would hold the power to correct domestic or supranational legislation, whenever it falls foul of CETA or TTIP. Yet, this outcome is no foregone conclusion, since the ECJ opts for a direct applicability of international agreements ‘only where the nature and the broad logic of the latter do not preclude this.’ In deciding whether this is the case, the Court considers, among other things, the purpose of the agreement, the will of the parties and the question of reciprocity, i.e. whether the Union would be alone in recognising direct effect. Luxembourg may have largely ignored the question of reciprocity with regard to Turkey and other neighbours of the EU, but it traditionally takes centre stage in the evaluation of trade liberalisation agreements.

Article 14.14 of the CETA Draft Treaty

Once we have understood the far-reaching implications of direct applicability, we may appreciate the bearing of the clause on private rights in the final provisions of the consolidated Draft CETA Agreement, which states: ‘Nothing in this Agreement shall be construed as conferring rights or imposing obligations on persons other than those created between the Parties under public international law, nor as permitting this Agreement to be directly invoked in the domestic legal systems of the Parties.’

That is nothing less than the official denial of direct effect in the operative treaty text; neither the ECJ nor domestic courts would hold the power to apply CETA rules in domestic proceedings or to annul legislation which contradicts trade law. CETA and TTIP would get the same treatment that the ECJ accords to WTO law, which also does not have direct effect in the EU legal order and the domestic legal systems of the Member States – not even in situations, in which the appellate body of the WTO Dispute Settlement Mechanism confirmed that EU legislation falls foul of WTO standards.

Denial of direct effect is an important stumbling block for the long-term success of any free trade agreement, especially when it comes to the elimination of non-tariff barriers to trade, since corresponding rules are often formulated in an open manner. The real-life implications of such vaguely formulated provisions depends decisively upon the continued will of the parties and the availability of control and enforcement mechanisms – as the experience with non-tariff barriers to trade in the EU single market and the example of language requirements for spouses of Turkish nationals demonstrate. Without institutional control mechanism, vague treaty formulations are often ineffective.

Implications for the Debate about Investor-State-Dispute Settlement

In the light of Article 14.14 of the CETA Draft Agreement, we may have to re-consider our perspective on the proposed investor-state-dispute settlement rules in both CETA and TTIP. Critics will use the absence of direct application as an argument to decry the special treatment for investors; supporters, by contrast, will argue with the Commission that the dispute settlement bodies are a compensation for domestic legal remedies, which would not be available for the substantive rules of CETA and TTIP. I personally share the opinion of Christian Tams that the debate should focus on the desirability of special rules for investors (and not the question of procedure). The latter may be superfluous in relations with the US and Canada, but to abandon them would render it difficult to insist upon similar provisions in negotiations with China, Russia or other states we trust less.

In any case, the absence of direct effect in domestic proceedings shows that the legal implications of CETA and TIIP would be less dramatic than some commentators in the public debate suggest. Courts in Europe would not hold the power to annul domestic or supranational legislation, which violates the agreements – a power that also the dispute settlement bodies would be denied explicitly, together with the option for the state parties to force their reading of the agreements upon the arbiters by means of an authoritative interpretation. This would orientate the dispute settlement rules in CETA and TTIP towards inter-state bargaining, pretty much like in the WTO context.

The denial of direct application demonstrates that both the CETA Draft Agreement and the TTIP negotiation position are far removed from resembling anything like EU-style integration through law – and I consider this to be a good thing given that any dynamic interpretation would lack much of the democratic legitimacy and procedural constraints, which we have, notwithstanding all the deficits, in the European context. It seems to me that it is the biggest benefit of the debate about CETA and TTIP to this date that the broader public has started discussing the governance of economic globalisation. That debate will stay with us, even if an seemingly unimpressive provision on page 470 of the CETA Draft Agreement excludes the domestic enforcement of the transatlantic trade rules.

This blogpost has been published previously on ‚Verfassungsblog – On Matters Constitutional‘

 

Barnard & Peers: chapter 25

Image credit: capreform.eu

Read More

In Vino Veritas: the CJEU again strengthens the EU’s external role

Steve Peers

Only a few Member States produce good wine; but it is quaffed enthusiastically in all of them. That simple fact lies at the heart of today’s judgment in Germany v Council, in which the CJEU, consistently with a string of recent judgments, significantly strengthened the EU’s role in external relations.

Background

In particular, this case concerned the situation where Member States are ‘trustees’ of the EU’s external competence. This occurs where (as is often the case), the EU is not able to sign up to a treaty or participate in an international organisation, even though it has external competence as regards the subject-matter of that organisation or treaty. In that case, as established in prior case law, Member States must act on the EU’s behalf.

However, the Treaties don’t regulate this situation directly, and before today, there was little case law regulating the details of such ‘trusteeship’. So the exercise of the EU’s powers in such scenarios can be awkward. For instance, earlier this year the Council could not agree on a Commission proposal to coordinate Member States’ positions on behalf of the EU as regards possible new ILO measures regarding forced labour (see further my blog post on this issue).

Today’s judgment concerned the International Organisation of Vine and Wine (OIV), an international organisation which was originally set up back in 1924, but which was reconstituted with a new name in 2001. Its membership includes 21 Member States, including some (such as Finland) which are not known for their quality of wine production, although the scope of the OIV also extends to grapes and (unfortunately for small children) raisins.

In practice, the main focus of the OIV is drafting technical resolutions on wine production and quality. These resolutions don’t bind the members as such, so essentially constitute ‘soft law’. The OIV adopts these measures at a lengthy annual conference held each summer in a rather pleasant location. There are probably many opportunities for delegates to sample all the latest fine wines. In short, OIV membership appears to be a little-known – but doubtless much-loved – perk for agriculture ministers and their officials.

But after 90 years of such genteel wine-tasting, this classy organisation was dragged into the mud of an arcane EU inter-institutional dispute. This began when the EU legislation establishing a common market organisation for wine was amended in 2008, in order to make the organisation’s resolutions binding as regards EU law. The Commission promptly proposed that the EU become a member of the organisation, alongside its Member States.

There was not enough support among Member States for the Council to adopt this proposal. Instead, the individual Member States coordinated their voting in the OIV’s annual meeting. The Commission then threatened those Member States with infringement actions if they continued to vote in the OIV to adopt measures which affected the EU acquis. So as a compromise, the Council agreed to adopt a Decision which would coordinate their position on behalf of the EU as a whole.

The Council acted on the basis of Article 218(9) TFEU, which provides as follows:

The Council, on a proposal from the Commission or the High Representative of the Union for Foreign Affairs and Security Policy, shall adopt a decision suspending application of an agreement and establishing the positions to be adopted on the Union’s behalf in a body set up by an agreement, when that body is called upon to adopt acts having legal effects, with the exception of acts supplementing or amending the institutional framework of the agreement

Germany, supported by several other Member States, then challenged this Decision on two grounds. First, it argued that Article 218(9) TFEU only permits the Council to establish the position of the EU as such, as distinct from the action of its Member States. Secondly, it argued that Article 218(9) TFEU only applies where the measures in question are binding as a matter of international law, rather than as a matter of EU law.

The judgment

The Advocate-General’s opinion agreed with Germany on both points, but the Court’s judgment rejected them both. First of all, as regards the scope of Article 218(9), the Court pointed out that the words ‘on the Union’s behalf’ did not specifically limit themselves to cases where the Union was party to the agreement in question. The Court did agree that the previous provisions of the external relations Title of the Treaty only referred to treaties signed by the EU. So did the reference to suspending a treaty, set out in Article 218(9) itself.

However, when it came to a decision adopted by a body established by an agreement, the word ‘agreement’ had a different, wider meaning, applying not only to treaties which the EU was a party to, but also to treaties which only the Member States (or some of them) were party to. The Court justified this distinction on the grounds that the Treaty rules on the negotiation, conclusion or suspension of agreements only applied to the EU itself, whereas the EU does not have to be a party to an agreement in order to control Member States’ actions as regards implementing measures.

On the second point, the Court described the soft-law decision-making process of the OIV, and then noted that due to the link with OIV measures made by EU law, those measures ‘were capable of decisively influencing the content’ of EU legislation. Therefore those measures had ‘legal effect’, and the Council could rightly adopt a Decision establishing the EU’s position on what they should be.  

Comments

The Court’s judgment means that the EU is in a strong position to coordinate its Member States’ action in international organisations, where the Member States are trustees of the Union’s external competence. In fact, as the Advocate-General’s opinion pointed out, there is now no real difference between the EU being a party to the OIV and not being a party to it.

First of all, there can now be no doubt that the EU has such a coordination power, on the basis of Article 218(9) TFEU. Secondly, since the scope of that power is linked to the existence and nature of the EU’s external competences, the Court’s recent judgment in the broadcasting rights case, giving a generous interpretation of the EU’s exclusive external powers after the Treaty of Lisbon, is relevant here.

That recent judgment is arguably also relevant by analogy to the interpretation of the condition that the planned acts must have ‘legal effects’. This condition applies, as today’s judgment makes clear, whenever EU legislation has made an express link to the measures which an international organisation might adopt. But the Court did not limit the notion of ‘legal effects’ to such cases. The concept might therefore also apply where there is merely an indirect potential impact upon EU legislation (cf the broadcasting rights case), or where the measure in question has effect only upon third parties, or within the international organisation itself.

What does this mean in practice? Where the EU has exclusive external powers, Member States can’t act to affect EU law, even if the EU is not a member of the international organisation (see the Commission v Greecejudgment, as regards the IMO). The Council Decisions coordinating Member State positions have to include a substantive legal base, so the nature of the EU’s competence is clear (CITES judgment). Even where the EU and the Member States share competence, and the EU has not regulated the issue in question, the existence of an EU strategy might prevent Member States from acting alone (see by analogy Commission v Sweden, although that case concerned a treaty which both the EU and its Member States were parties to).

Of course, the obligation of Member States to act as trustees in the EU’s interest does not mean that they will necessarily agree on a proposed Council Decision to coordinate their action (for instance, see the example of the ILO forced labour measures, referred to above).

The Court’s judgment necessarily does not directly touch upon the question of the process by which the EU authorises its Member States to sign or conclude treaties (as distinct from acting within an international organisation) as trustees of EU competence. But if the judgment is read literally, it will have a significant impact on that process. For although the judgment states that Article 218(1) to (8) TFEU only applies to the negotiation and conclusion of agreements by the EU, the EU practice until now is to use these provisions also to approve the negotiation and conclusion of treaties by the Member States, acting as trustees of the EU’s interest. (See for instance, the Council Decision concerning the domestic workers convention, based on Article 218(6) and (8)).

Oddly, there was no real discussion in this litigation of the possible alternative route of using the EU’s internal legislative powers to regulate Member States’ behaviour within international organisations. The Advocate-General briefly (and bizarrely) mentioned the possible use of Article 352 TFEU, the ‘residual powers’ clause in the Treaties, but why not simply use Article 43 TFEU, the power relating to the common agricultural policy?

This would entail the adoption of a legislative act. While it might be argued that this is too cumbersome a process to use every time the OIV holds an annual meeting, it might instead be possible to adopt only one legislative measure, which sets out a general framework for coordinating Member States’ action as regards all future OIV meetings.

Adopting a legislative act in relation to an international treaty might seem odd at first sight, but it isn’t really. It was, of course, a legislative act that first gave legal effect to OIV soft law in the EU legal order in the first place. More broadly, the EU often adopts legislative acts to coordinate Member States’ treaty-making competence, in areas subject either to EU exclusive competence (cf investment agreements) or to shared competence, where there is a large EU role (cf air transport treaties).

And if the Court really meant to say that Article 218(1) to (8) can no longer be used to allow Member States to sign and ratify international treaties as trustees of EU competence, then legislative acts will have to be used in this context too. So it would have made more sense for the Court to rule that EU internal legislative powers must be used to regulate all aspects of Member States’ trusteeship.

Some final thoughts on the role of the other EU institutions, and the position of non-participating Member States, following this judgment. First, the European Parliament (EP). It didn’t participate in the proceedings, but perhaps it should have done. As the Advocate-General pointed out, the judgment is the worst-case outcome for the EP, since it did not have the opportunity to approve EU membership in the OIV, and nor can it control the Council’s adoption of measures which impact ultimately upon the interpretation of EU legislation. In future, the EP would have the opportunity to address such issues when the EU legislation making such a link to international measures is adopted. But in this case, the legislation was adopted before the Treaty of Lisbon, when the EP was only consulted upon agricultural legislation. At least, as the Advocate-General pointed out, the EP must be ‘immediately and fully informed’ of the Council Decisions relating to Member States’ trusteeship pursuant to Article 218(10) TFEU, which the CJEU has recently interpreted broadly.

As for the CJEU, it has special jurisdiction relating to envisaged international treaties pursuant to Article 218(11) TFEU. If Article 218(9) applies to Member States’ trusteeship, then surely so does Article 218(11). Indeed, as the Advocate-General pointed out, the Court has already ruled as much. In fact, it will imminently be ruling again on such a case (Opinion 1/13, on the Hague Convention on child abduction). So we will be able in a week’s time to see how the two judgments fit together.

Finally, what about the position of Member States which are not members of the OIV? Actually, the operative part of the Court’s ruling makes no reference to this issue, and the judgment is logically equally applicable whether some or all Member States are party to the international agreement in question.

But in the particular context of the OIV, the Court’s confirmation that the Council has the power to coordinate Member States’ positions in the EU’s interests makes obvious sense. Due to the link between OIV resolutions and EU law, those resolutions impact all Member States, because they affect the quality and price of wine drunk in every Member State. Furthermore, they affect the operation of the common market in wine, which is paid for by all EU taxpayers, whether they are teetotallers or exclusively drink beer or even (inexplicably) whisky.  In this light, the decision of the UK (not a member of the OIV) to intervene, with other Member States, in support of Germany, rather than the EU institutions, is simply Pavlovian.

Barnard & Peers: chapter 24

Read More