It is the question of Member State involvement, not the technicalities of EU foreign relations law, which explains the interest of the wider public in the ECJ’s Opinion 2/15. To what extent will the EU institutions be allowed to act without the involvement of national parliaments when concluding free trade agreements with Singapore, Canada, the US or, potentially, China? It was widely acknowledged, not least in the excellent post by Daniel Sarmiento, that judges in Luxembourg delivered a smart compromise formula granting the EU institutions much leeway, while assuring that the Member States control the future of Investor-State Dispute Settlement (ISDS). Yet, initial comments also claimed that the Court seemed to have eliminated the option of ‘facultative’ EU-only agreements which do not embrace ISDS (here and here). It seems to me that this overlooks an important novelty of Opinion 2/15, which gave explicit judicial blessing to the option of ‘facultative’ EU-only agreements, although the Court hides the innovation behind an inconsistent use of the notion of ‘shared’ powers.
The Concept of ‘Facultative’ EU-only Agreements
Much of the debate on the future of mixity rightly focused on the scope of the revised Treaty provision on the Common Commercial Policy (CCP) that was interpreted broadly by the ECJ in Opinion 2/15. However, EU external action is not about trade alone: it can encompass many other fields for which different legal bases are relevant. Corresponding competences can be laid down in express Treaty provisions, such as Article 208 TFEU on development cooperation or Article 191(4) TFEU on the environment, but they can also flow from the general system of the EU Treaties. These competences have traditionally been called ‘implied’ powers following the rationale of the seminal AETR (or ERTA) judgment of 1972. These implied powers can take over when explicit legal bases, such as the Treaty provision on the CCP, do not cover an area. The most prominent example are transport services, which are excluded from the CCP by virtue of Article 207(5) TFEU and in relation to which the ECJ recognised in Opinion 2/15 that they are subject to an (exclusive) AETR competence (paras 168-217). It is on such implied powers that the Opinion presents a decisive novelty.
In practice, implied powers have so far became relevant mainly in situations when the EU institutions claimed an exclusive competence after the adoption of internal legislation. That is why the discussion on implied powers is often associated with exclusivity, although other scenarios are similarly valid: many contend that the EU can activate an implied external power by concluding an international agreement. It would not be obliged to do so, but it could decide politically to have recourse to its shared powers. Such a conclusion would not constitute an ultra vires act provided that the EU’s external powers flow from a sufficiently clear legal basis – a requirement the original AETR case law satisfies in the eyes of the German Constitutional Court, which declared the concept of implied powers to be methodologically sound in light of public international law and US constitutional doctrine (see here, para 237). Moreover, implied powers have been codified by the Treaty of Lisbon and are, as a result, not ‘implied’ in a strict sense any more. They can be found in Article 3(2) TFEU on exclusivity in Article 216(1) TFEU on international agreements more generally – a distinction Opinion 2/15 builds upon.
How can such implied non-exclusive power be exercised? The answer is straightforward: through an activation during the Treaty conclusion procedure in accordance with Article 218 TFEU. For reasons of legal certainty and parliamentary oversight, a corresponding desire should be stated explicitly by the institutions, for instance when authorising the signature or conclusion of an agreement. If the above is correct, the EU can conclude ‘facultative’ EU-only agreements provided that they do not contain any provisions requiring Member State involvement beyond the reach of exclusive or shared EU competences. An example may be the recent Stabilisation and Association Agreement with Kosovo, which covers various fields that have traditionally been considered to require a mixed agreement, while the said treaty was signed and ratified by the EU (and the EAC) on the basis of Article 217 TFEU, thereby streamlining ratification requirements after the negative Dutch referendum on the association agreement with Ukraine. Could this model be replicated in relations with neighbouring countries and for trade deals, provided the latter do not include provisions on ISDS? It seems to me that Opinion 2/15 moved decisively in that direction.
Non-Exclusive Shared Competences in Opinion 2/15
The key argument is part of the section on non-direct portfolio investments, which the Court considered not to be covered by the CCP (thereby reaffirming the critical position of the German Constitutional Court in its Lisbon judgment). To exclude non-direct foreign investments from the scope of the CCP does not mean, however, that the EU cannot act. We need to consider, rather, whether the area is covered by an implied power (something the German Constitutional Court failed to discuss in paras 52-57 when rejecting a preliminary injunction against the CETA with Canada). That is what the ECJ did in paras 225-257 of Opinion 2/15.
Much of the debate on portfolio investments revolved around the existence of an exclusive implied power in line with the argument brought forward by the Commission. The ECJ rejected this position, rightly in my view, as the sole existence of an EU competence, which – in contrast to the field of transport – has not been exercised, cannot justify the Commission’s standpoint on exclusivity (paras 229-235). Crucially, judges did not stop at the rejection of an exclusive Union power, but proceeded with an analysis whether non-direct foreign investments constitute a shared competence. In doing so, the Court had recourse to the (new) Treaty provision in Article 216(1) TFEU whose second alternative allows the EU to conclude an agreement ‘where [this] is necessary in order to achieve, within the framework of the Union’s policies, one of the objectives referred to in the Treaties.’ It considered that rule to apply to portfolio investments (paras 239-240) and concluded, in light of Article 4(6) TFEU, that the competence is shared (paras 241-242).
That argument appears to be straightforward: the ECJ’s explicit reference to Article 216(1) TFEU in conjunction with Article 2(6) TFEU indicates rather unambiguously that it employs the term ‘shared competence’ within the meaning of Articles 2-6 TFEU, i.e. the EU can conclude agreements on portfolio investments in line with Article 2(2) TFEU, but the competence is non-exclusive (in contrast to the Commission’s claim). Such recourse to Article 216(1) TFEU for the identification of a non-exclusive implied power constitutes a major novelty in doctrinal terms, since it distinguishes between exclusive implied powers after the adoption of internal legislation (covered by Article 3(2) TFEU on which the Commission had relied) from political discretion about whether to exercise non-exclusive shared competences (dealt with in Article 216(1) TFEU). This distinction rationalises in retrospect the rather cumbersome codification of the implied powers doctrine in the Treaty of Lisbon discussed by Marise Cremona. It was not the first time, the Court took seriously the new provisions on external action: in Opinion 2/15, it had recourse to the new Treaty objectives to extend the CCP to sustainable development; in an earlier judgment, it used Article 16-17 TEU to delineate the prerogatives of the Commission and the Council.
Judicial Confusion over the Meaning of ‘Shared’ Competence
Unfortunately, what appears to be a clear-cut argument on the basis of Article 216(1) TFEU read in conjunction with Article 2(6) TFEU is rendered more complex by the conclusion, on the part of the ECJ, that rules on portfolio investments ‘cannot be approved by the European Union alone’ (para 244). This takes up a different reading of ‘shared’ competence, which contradicts the constitutional rationale of Article 2 TFEU, cited by the ECJ just before: the alternative vision refers to mandatory Member State involvement as a result of limited EU powers. It is employed similarly, in Opinion 2/15, with regard to ISDS, which the Court considered generally not to transcend EU competences, since it did not mention – in contrast to portfolio investments – any legal basis in the EU Treaties (paras 285-293), i.e. ISDS seems to be covered by neither an exclusive nor a shared competence in the meaning of Article 2 TFEU. Nonetheless, the ECJ concluded that ISDS is subject to a ‘competence shared’ (French: compétence partagée) between the EU and the Member States and that, therefore, both have to consent to the agreement.
This alternative use of the notion of ‘shared competence’ as an indicator for mandatory mixity due to limited EU powers is no novelty. It was employed by many of the Court’s seminal external relations cases, including Opinion 1/94 on a ‘shared’ competence for the GATS (para 98) and Opinion 1/03 in which the Court defended a binary distinction between exclusive Union powers and mixed (shared) agreements (para 115). Opinion 2/15 builds on this semantic tradition, although it contradicts the constitutional concept of ‘shared competence’ introduced by the Treaty of Lisbon following the debate prior to the Constitutional Treaty. In the Court’s conclusion on portfolio investments, the semantic path-dependency of judicial formulae and the novel Treaty articles on Union competences are in open conflict.
How can we disentangle this outcome? One option is to distinguish analytically between the use of the term ‘shared’ for the purposes of Article 216(1), 4(2) TFEU and the designation of a mixed ratification requirement. In the first case, the term ‘shared’ is employed as an adjective describing the character of the competence; in the second case, it is used as a participle defining the overall situation. While the English and the French language versions of Opinion 2/15 support this distinction, the German language version, amongst others, employs identical language as Article 4 TFEU despite the conclusion that the EU cannot act alone. Even if we managed to disentangle the semantics, the Court’s conclusion on portfolio investments would remain contradictory if, on the one hand, the EU has a non-exclusive shared competence on the basis of Article 216(1) TFEU, which, on the other hand, it cannot exercise alone without Member State involvement. In short, Opinion 2/15 is inconsistent, but it seems to me that the above-mentioned innovation on Article 216(1) in conjunction with Article 4(2) TFEU is so explicit that it will be difficult to discard the concept in the future.
Scope of Non-Exlusive Implied Powers
The contours of the EU’s shared power under Article 216(1) TFEU are more than a technicality, since the scope of non-exclusive implied powers would define the constitutional limits of any political decision to move towards ‘facultative’ EU-only agreements. Indeed, it is apparent that doing so would require a political decision in line with the underlying idea of shared powers under Article 2(2), (6) TFEU explicitly referred to by the ECJ. Thus, the Commission could not oblige the Council to act; the latter would have discretion during the procedure prescribed in Article 218 TFEU, which also provides, in most scenarios, for the consent of the EP. Doctrinally, it remains a political decision whether the conclusion of international agreement is ‘necessary’ in the meaning of Article 216(1) TFEU. That is an important difference to the idea of mandatory exclusivity without the prior adoption of internal legislation which the ‘necessity’ requirement under the second alternative of Article 3(2) TFEU provides for in line with the exceptional situation that gave rise to Opinion 1/76 on the regulation of inland waterways in the wider Rhine area. By contrast, the second alternative of Article 216(1) TFEU is a question of political choice on the side of the Council, which will regularly act by a qualified majority under Article 218(8) TFEU.
Besides the procedural requirements, the Council will have to respect the substantive limits of Article 216(1) TFEU, which, as an attributed power, is subject to limitations (Article 5(1) TEU). In this respect, the wording of the provision and Opinion 2/15 can be read in two alternative ways. Firstly, it could be argued that Article 216(1) mirrors Article 352 TFEU with its reference to the realisation of Treaty objectives, thereby authorising the adoption of secondary legislation also in the absence of an explicit competence in the EU Treaties. That is what the ECJ seems to suggest with regard to portfolio investments in Opinion 2/15, when it postulates a shared power on the basis of Article 216(1) read in conjunction with Article 63 TFEU (para 240). It is not clear, given the brevity of the reasoning, whether this lends support to a shared competence, on the basis of Treaty objectives, which is broader than the competence for the adoption of secondary legislation under Article 64 TFEU. If that was the case, such conclusion might be scrutinised critically by the German Constitutional Court, which famously equated Article 352 TFEU with a carte blanche in the grey area between an express power and an ultra vires act (see the Lisbon judgment, paras 352-328).
Secondly, it would be possible to limit the reach of the second alternative of Article 216(1) TFEU to the scope of existing internal powers. This interpretation could build on the wording, which refers to the ‘framework of the Union’s policies’, although it should be acknowledged that Article 352(1) TFEU says the same. More importantly, however, such parallelism would reiterate the original rationale underlying the implied powers doctrine that the EU competence to act internationally mirrors its ability to adopt internal legislation – or, in Latin: in forum interno, in forum externo. For the purposes of future free trade agreements, this would entail that not all rules on portfolio investments would covered by a non-exclusive shared competence automatically, since the reach of the EU’s internal powers under Article 64 TFEU and other provisions would have to be ascertained. Possibly, it was this uncertainty about the scope of the EU’s shared competence on portfolio investments, which motivated the ECJ to evade a clear-cut conclusion on that matter in Opinion 2/15. It will have the opportunity to clarify its position, thereby reaffirming the innovative potential (and the limits) of the novel non-exclusive implied powers under Article 216(1) TFEU