Introduction: the context and significance of the Opinion

On 16 May 2017 Opinion 2/15 was delivered by the full Court, in response to a request by the Commission pursuant to Article 218(11) TFEU.Footnote ¹ According to that provision, which is found in the Article dealing with the negotiation and conclusion of international agreements by the European Union, ‘[a] Member State, the European Parliament, the Council or the Commission may obtain the opinion of the Court of Justice as to whether an agreement envisaged is compatible with the Treaties’. Although the request is for an ‘opinion’, Article 218(11) TFEU also provides that ‘[w]here the opinion of the Court is adverse, the agreement envisaged may not enter into force unless it is amended or the Treaties are revised’; it therefore has a binding legal effect.

The ‘agreement envisaged’ in this case was the free trade agreement that had been negotiated between the EU and Singapore (‘the Singapore Agreement’). The gestation of this agreement had been long: the original authorisation to negotiate had been granted by the Council in 2006 and envisaged a region-to-region agreement with ASEAN as a group.Footnote ² However the Council also envisaged the possibility of reverting to bilateral agreements with individual ASEAN states, and this is in fact what happened: the ASEAN negotiations were paused and on 22 December 2009 – as the Court points out, a few weeks after the Lisbon Treaty came into force – the Council authorised the Commission to start negotiations with Singapore.Footnote ³ In 2011 the negotiating directives were amended to include investment and in June 2015 the text was initialled. The agreed text assumes that the Singapore Agreement will be concluded by the EU alone, rather than by the EU and its member states as a ‘mixed’ agreement.Footnote ⁴ The Council and member states were, however, not convinced that – as the Commission argued – the whole agreement fell within exclusive EU competence; they thought the agreement contained elements of shared competence, and indeed elements falling within exclusive member state competence, and that it should therefore be mixed.

Similar discussions were taking place over other negotiations as the EU sought to conclude a new generation of trade and investment agreements with (among others) Canada and the USA. In fact, the Singapore agreement is an example – and not the most controversial – of these new ‘deep and comprehensive’ trade agreements that represent the core of EU trade policy, and the Court’s Opinion is thus important not only for the future of the Singapore Agreement itself, but for future trade agreements (including a possible future EU-UK agreement). The current wave of negotiations date from the ‘Global Europe’ Trade Strategy adopted in 2006,Footnote ⁵ when the EU decided to reinvigorate its programme of trade liberalisation based on bilateral agreements with important trading partners. This was a response both to failures to progress multilateral negotiations within the World Trade Organisation and to the decision by the USA to pursue major regional free trade agreements, especially in the Asia-Pacific region.Footnote ⁶ In addition to the traditional liberalisation of trade in goods these recent agreements aim to extend World Trade Organisation commitments on trade in services and enforcement of intellectual property rights, and to improve regulatory cooperation. They have also featured chapters on trade and sustainable development, which link trade liberalisation to commitments to internationally-agreed goals and standards of environmental protection, renewable energy and core labour rights.Footnote ⁷ Since the coming into force of the Lisbon Treaty, the Union has also added investment, both market access and investor protection, and in some cases (including the Singapore Agreement) investor-state dispute settlement (ISDS). The potential impact of investor protection and ISDS on the regulatory autonomy of the EU and its member states has proved controversial and has given rise to highly visible civil society campaignsFootnote ⁸ and political debate.Footnote ⁹ Alongside the political debates, legal debates have taken place over the compatibility with EU law of the provision in the recent agreements for ISDS either, traditionally, via arbitration (as in the Singapore Agreement) or via the establishment of an investment court (as in the agreements with Vietnam and Canada).

Legal debate has also centred on the distribution of competence to conclude these new extensive trade agreements. This involves the extent of the post-Lisbon common commercial policy (CCP), a matter of a priori exclusive competence as a result of Article 3(1) TFEU,Footnote ¹⁰ and in particular the extent to which the CCP covers investment protection; but also concerns the consequences of including chapters on trade and sustainable development: could these chapters be subsumed under CCP powers or did they need a separate legal basis? This group of questions over competence are at the heart of Opinion 2/15.Footnote ¹¹ The Commission’s request was restricted to the question of competence, and conspicuously omitted any question as to substantive compatibility:

1. 1. Which provisions of the agreement fall within the Union’s exclusive competence?

2. 2. Which provisions of the agreement fall within the Union’s shared competence? and

3. 3. Is there any provision of the agreement that falls within the exclusive competence of the Member States?Footnote ¹²

That these questions were regarded as important, not only for the Singapore Agreement itself, but also as setting a precedent for the negotiation and conclusion of the next generation of EU trade agreements, can be seen from the fact that 25 Member States submitted observations.Footnote ¹³ In its Opinion the Court (and in her opinion, the Advocate GeneralFootnote ¹⁴ ) focuses strictly on the question of competence. Three times the Court repeats that its Opinion in this case is without prejudice to the issue of compatibility;Footnote ¹⁵ it cannot therefore be assumed that the absence of discussion is a sign that the Court considers that there are no compatibility problems – perhaps indeed rather the contrary. That the Court was well aware that compatibility has been debated, in particular with respect to the provisions on ISDS, is clear from the fact that one of its reminders that its findings are without prejudice to compatibility occurs in the passage of the Opinion which discusses competence in relation to ISDS.Footnote ¹⁶ And indeed, since the handing down of Opinion 2/15, a new request under Article 218(11) TFEU has been made, this time by Belgium, with respect to the EU-Canada Comprehensive Economic and Trade Agreement, asking whether the ISDS provisions of the Agreement are ‘compatible with the Treaties, including fundamental rights’.Footnote ¹⁷

The following comments on Opinion 2/15 will therefore not discuss the compatibility of ISDS with the EU Treaties,Footnote ¹⁸ but it should be borne in mind that the Opinion does not settle all legal questions arising from this latest generation of trade and investment agreements. Nevertheless, it is an important Opinion. The increased political salience of trade agreements in a world of anti-globalisation and challenges to the post-war Western consensus in favour of trade liberalisation means that the issue of who ultimately gets to decide on EU trade policy, and especially the conclusion of trade agreements, is no longer a purely academic or technical question. Indeed although the Court has always taken the view that procedural (or political) complications should not drive the determination of competence,Footnote ¹⁹ it is clear that the design of trade agreements is being shaped by the implications of including chapters which fall within shared competence and where Member State participation is very likely,Footnote ²⁰ even if not legally required, with all that means in terms of the political uncertainties of ratification.Footnote ²¹ Opinion 2/15 was thus eagerly awaited, as the Court’s determination of the boundaries of exclusive EU competence will certainly be a factor in deciding on the content of future trade agreements, and especially the inclusion of investment protection and ISDS. More prosaically perhaps, Opinion 2/15 forms an important part of the Court’s interpretation of the Lisbon Treaty’s codification of EU law on external competence; not only the interpretation of the (enlarged) scope of the CCP but also the application of Article 3(2) TFEU on exclusivity as regards external agreements, and the basis on which external powers may be derived from internal competence-conferring provisions. In this it forms part of a broader picture and indeed case law since Opinion 2/15 was delivered has already added to, and nuanced, the picture.Footnote ²²

The Opinion: a summary

The Opinion is long and complex, and the opinion of Advocate General Sharpston even longer, at 570 paragraphs, with an Annex containing a 131-paragraph summary of the Singapore Agreement. In this note I highlight those parts of the Advocate General’s opinion where she takes a different view from that of the Court, but the opinion as a whole repays close reading; in particular, the first part of the opinion offers a clear and succinct summary of the relationship between internal and external competence and legal basis, the conditions under which exclusive competence arises, and the ‘state of the art’ in understanding them in the light of pre- and post-Lisbon case law.Footnote ²³ In the following sections I will comment on some key aspects of the ruling: the scope of the EU’s common commercial policy and the role played by the general external objectives established in the EU Treaties; the application of the conditions for exclusive external competence set out in Article 3(2) TFEU; whether shared competence necessarily implies mixity (the conclusion of the agreement by the member states and EU jointly); the question of the member states’ bilateral investment treaties; and competence in relation to ISDS. But first I will give a general orientation by briefly summarising the Court’s Opinion.

The Advocate General and the Court came to the same conclusion in fine: that the Singapore Agreement does not fall wholly within the EU’s exclusive competence. They also agreed on a central question: that non-direct foreign investment, including portfolio investment, falls outside the CCP and is currently a matter of shared competence. However there are also significant differences, of approach and in their conclusions. The Advocate General found greater areas of shared competence in the agreement, and even one clause (on the termination of Member State bilateral investment treaties) that in her view was not within EU competence at all, but was an exclusive Member State competence. The Court of Justice held that all aspects of the agreement fell within exclusive EU competence – on the basis either of Article 3(1)(e) TFEU (the CCP) or of Article 3(2) TFEU – except for (i) the provisions on investment protection insofar as they relate to non-direct investment; (ii) institutional and dispute-settlement provisions insofar as they relate to non-direct foreign investment; and (iii) the provisions on ISDS.

Building upon its earlier post-Lisbon judgments on the scope of the CCP,Footnote ²⁴ in Opinion 2/15 the Court was willing to find a sufficiently close link to the agreement’s trade objectives for a variety of clauses – including, crucially, the chapter on sustainable development – to be encompassed by the CCP. In this the Court was more explicit than it had been before as to the impact of the general external objectives on the nature of the Union’s trade policy. The Court also gave a relatively expansive application of the test for exclusivity found in Article 3(2) TFEU, more specifically the final limb of that provision: that the (provision in the) agreement ‘may affect common rules or alter their scope’, finding this condition satisfied more easily than the Advocate General. As a result a wide-ranging trade and investment agreement was found to fall within exclusive EU competence – all except certain specific provisions on non-direct investment protection and on ISDS. It is not surprising that a number of commentators have suggested that the Court is effectively inviting the Commission to alter tactic, removing provisions on investment protection (or non-direct investment protection) and ISDS from trade agreements and placing them into separate agreements.Footnote ²⁵ The main trade agreement could then be concluded under exclusive powers; the (politically more controversial) investment protection/ISDS agreement would need to be concluded as a mixed agreement and would, as well as taking much longer to ratify, require explicit political commitment from each Member State.

It is notable, nevertheless, that the Court itself makes no such suggestions. It does not engage even tangentially with the practical issues surrounding its conclusion that the Singapore Agreement falls within shared competence and the implications of mixity.Footnote ²⁶ In fact, as has been pointed out by Kleimann and Kübek,Footnote ²⁷ the Court is even somewhat coy in its final conclusions: it concludes that aspects of the agreement fall within exclusive EU competence, and other aspects within shared competence, but it does not in terms answer the Commission’s main question: does the EU have the requisite competence to conclude this agreement alone? Whether mixity is indeed required (as opposed to an option) in cases of shared competence is a question raised by the Opinion, and will be discussed further below; but it is at least clear that the Court was aware that the choice of the member states was in favour of mixity for these politically significant agreements.Footnote ²⁸ In the Advocate General’s opinion mixity was required because one clause in her view fell within exclusive Member State competence; she alludes to the difficulties of mixed agreements and to the options available, including splitting the agreement, but as she rightly says, these are not factors which can or should influence the determination of competence.Footnote ²⁹

Let us now turn to the main issues in the case.

The scope of the common commercial policy

At the centre of this case was the scope of the post-Lisbon common commercial policy, and the extent to which this legal basis can encompass the wide-ranging provisions of a modern trade agreement.Footnote ³⁰ The Lisbon Treaty had brought within the scope of exclusive trade competence, trade in services, the ‘commercial aspects’ of intellectual property and foreign direct investment.Footnote ³¹ Earlier judgments had established the Court’s approach to interpreting the scope of the post-Lisbon CCP in relation to trade in servicesFootnote ³² and the commercial aspects of intellectual property.Footnote ³³ This approach has focused not on attempts to define these terms in the abstract, but rather on identifying the scope of the CCP as a policy concerned with international trade (i.e. trade with non-member states). CCP measures, the Court has said, must have a ‘specific link’ with international trade, in that they are intended to ‘promote, facilitate or govern’ trade and have ‘direct and immediate effects’ on trade.Footnote ³⁴ And in seeking to distinguish between measures falling within the CCP and those which should properly be based on other powers, such as the internal market, the Court has recognised that harmonisation of the laws of the member states is not a matter for the CCP; the CCP can, however, be used in order to extend to third countries standards currently applicable within the EU as a result of internal market legislation.Footnote ³⁵ This approach to the CCP underlies, and is reinforced by, Opinion 2/15. The Court summarises its earlier case law in a paragraph which it subsequently refers back to several times:

Applying this test to the provisions on market access, the Court had no difficulty in finding that chapters 2-6 of the Singapore Agreement, which concern trade in goods, as well as chapter 7 on non-tariff barriers to trade and investment in renewable energy generation and chapter 10 on public procurement, fell within the CCP.Footnote ³⁷ In relation to trade in services (chapter 8), the Court explicitly confirmed that all modes of supply of services fall within the CCP;Footnote ³⁸ and that the only remaining sectoral exclusion is that of agreements ‘in the field of transport’, which according to Article 207(5) TFEU (continuing an exception introduced by the Treaty of Nice) are governed by the Treaty provisions relating to the common transport policy.Footnote ³⁹ The provisions on transport services in chapter 8 of the Singapore Agreement do not therefore fall within CCP-based exclusive competence and were considered by the Court later in its Opinion, in considering exclusive competence on the basis of Article 3(2) TFEU.

In examining the Singapore Agreement’s chapter on intellectual property rights, the Court first affirms that the ‘commercial aspects’ of intellectual property in Article 207 TFEU refers to those aspects which satisfy the test of possessing a ‘specific link’ to international trade. The different provisions on intellectual property in the Singapore Agreement consist of ‘first, a reminder of existing multilateral international obligations and, secondly, bilateral commitments’.Footnote ⁴⁰ The purpose of these provisions, and those dealing with enforcement of intellectual property rights and counterfeit and pirated goods, is to guarantee entrepreneurs in both parties an ‘adequate level’ of protection of their rights, so as to ensure ‘a degree of homogeneity’ in standards of protection of intellectual property rights and enforcement.Footnote ⁴¹ This latter phrase allows the Court to argue that the intellectual property provisions are about ‘increas[ing] the benefits from trade and investment’ and liberalisation of trade,Footnote ⁴² but are ‘in no way’ about harmonisation.Footnote ⁴³

Chapter 9 of the Singapore Agreement on investment protection provided the first opportunity since the Lisbon Treaty reform for the Court to address the question of the application of the CCP to foreign direct investment. Here again the Court applies the ‘direct and immediate effects on trade’ test. In a first step, a distinction is made between foreign direct investment and other forms of foreign investment. Unsurprisingly, the Court found that the express qualification of ‘direct’ investment in the Treaties meant that the drafters had wished to exclude other – non-direct – foreign investment; it then turned to earlier case law interpreting ‘direct investment’ in the context of the movement of capital, to the effect that

This includes the acquisition of shareholdings in a company where the holding enables the shareholder ‘to participate effectively in the management of that company or in its control’.Footnote ⁴⁵ The Court here established a parallel between ‘direct investment’ in internal and external contexts. However, it then brings the external dimension into focus by saying that this type of direct investment meets its ‘direct and immediate effects on trade’ criteria. Indeed, this definition of foreign direct investment, although taken from an internal context, is justified on the ground that it reflects the scope of the CCP:

The same criterion is then used to argue that the CCP may cover all measures which impact foreign direct investment: not only those concerned with market access, but also those which concern investment protection (the subject of chapter 9 of the Singapore Agreement) including fair and equitable treatment and protection against expropriation without compensation. Arguments that only market access falls within the CCP are rejected.Footnote ⁴⁷ The provisions on investment protection ‘are such as to have direct and immediate effects on that trade, since they concern the treatment of the participation of entrepreneurs of one Party in the management or control of companies carrying out economic activities in the territory of the other Party’;Footnote ⁴⁸ each ‘contributes to the legal certainty of investors’ and meets the CCP test:

In places the investment protection measures refer to areas of national law such as public order, public security, taxation and criminal law. Although these were argued to fall within Member State powers, the Court held – in line with its own case law on property ownership, taxation and intellectual property in an internal context – that the Singapore Agreement’s requirements deal rather with how those powers are exercised. The provision in question

Likewise, the provision on expropriation does not impinge on Member State powers as regards property ownership (cf Article 345 TFEU); member states are free to exercise those competences, as long as they do so ‘in compliance with general principles and fundamental rights, in particular with the principle of non-discrimination’.Footnote ⁵¹

The Opinion therefore answers a number of questions on the scope of the CCP as regards foreign direct investment. The relatively inclusive approach is based on the ‘direct and immediate effects on trade’ test which is now well-entrenched and will be the starting point in future cases raising the scope of the CCP. Its mode of application by the Court in the Opinion gives an impression of inevitability of outcome: it is presented as obvious that some provisions have a ‘specific link’ to trade, while others do not. It is, however, a test which in fact allows the Court a great deal of room for manoeuvre in its application, as its reasoning in relation to investment and to intellectual property protection demonstrates.Footnote ⁵² That means that it will not necessarily be easy to take an objective view of how it should be – and would be – applied in a specific case; it is a test which favours flexibility over predictability. This may be pragmatic and desirable in a test which, as well as covering the existing very varied provisions in trade agreements on goods, services, intellectual property and foreign direct investment, will need to adapt to new types of trade agreement. But it does not necessarily provide certainty in future cases: we may have an answer for specific types of clauses in current trade agreements, but disputes as to the application of the test in different contexts are likely to end in further litigation.Footnote ⁵³ And the Court will need to be careful in presenting its reasoning so that over time a coherent picture is able to emerge.

Sustainable development and trade – the importance of general external objectives

We have so far examined those aspects of the Opinion where the Court draws the boundaries of what may legitimately fall substantively within the references in Article 207 TFEU to goods, services, intellectual property and foreign direct investment. But there is another equally important aspect to the extent of the CCP as a legal basis, which relates to the ability of a trade agreement to include provisions designed to achieve non-economic objectives. A classic approach would be to ask whether such provisions are simply ancillary or incidental to the predominant trade objectives or whether they impose independent obligations sufficiently substantive to constitute a distinct objective requiring a separate legal basis.Footnote ⁵⁴ In a clear departure from the opinion of the Advocate General the Court found that the provisions of chapter 13 of the Singapore Agreement, headed ‘trade and sustainable development’, also met the conditions of having a specific link to international trade, and direct and immediate effects on trade, and could indeed be seen as an integral part of trade policy. They could thus legitimately also be based on Article 207 TFEU, even though they do not appear to be trade instruments in the traditional sense. There are essentially two parts to the Court’s argument.

The first redefines the proper objectives of EU trade policy in the light of the general objectives and principles which are to govern all external action and the explicit injunction that trade policy must be conducted ‘in the context of’ these principles and objectives.Footnote ⁵⁵ Relying on Article 21 TEU (which includes sustainable development among the EU’s external objectives), Articles 205 and 207(1) TFEU as well as Articles 9 and 11 TFEU (referring to the integration of social protection and environmental protection requirements into all EU policies and activities ‘with a view to promoting sustainable development’), the Court concludes that ‘the objective of sustainable development henceforth forms an integral part of the common commercial policy’.Footnote ⁵⁶ This is a significant move, signalling that the Court is willing to take seriously the catalogue of general external objectives in Article 21 TEU as well as the reference to ‘free and fair trade’ in Article 3(5) TEU, as complementary to and capable of influencing the objective of trade liberalisation spelled out in Article 206 TFEU.Footnote ⁵⁷ As a result, the broader objectives of the Singapore Agreement could be subsumed under the umbrella of trade policy, without needing to be categorised as ancillary or incidental to the predominant purpose. The Advocate General took a different position; while recognising the importance of the general external objectives, she insisted that Articles 3(5) and 21 TEU and Articles 9 and 11 TFEU, although potentially affecting the compatibility of EU action with the Treaties, ‘cannot affect the scope of the common commercial policy laid down in Article 207 TFEU’ or ‘modify the scope of the European Union’s competence’.Footnote ⁵⁸

In a second step, the Court examined the different provisions in chapter 13 in the light of the established test for the CCP, based on the specific link to international trade. Here again it took a different view from the Advocate General. Advocate General Sharpston had concluded that whereas some provisions in chapter 13 (such as those promoting climate-friendly goods and services) had a direct link to international trade, others are in themselves standard-setting, albeit by reference to other treaties such as ILO Conventions and multilateral environmental agreements which both the EU and Singapore have ratified. In her view these commitments ‘essentially seek to achieve in the European Union and Singapore minimum standards of (respectively) labour protection and environmental protection, in isolation from their possible effects on trade. Those provisions therefore clearly fall outside the common commercial policy’.Footnote ⁵⁹

The Court, following the line of reasoning it had applied to intellectual property conventions, took a different view of these commitments. Far from being separate from the trading relationship, they are part of it; indeed, chapter 13 ‘plays an essential role in the envisaged agreement’.Footnote ⁶⁰ The parties ‘undertake, essentially, to ensure that trade between them takes place in compliance with the obligations that stem from the international agreements concerning social protection of workers and environmental protection to which they are party.’Footnote ⁶¹ These references do not, it held, turn the commitments they contain into new commitments under the Singapore Agreement: ‘Chapter 13 concerns neither the scope of the international agreements to which it refers nor the competences of the European Union or the member states relating to those agreements’.Footnote ⁶² They do not therefore require a separate legal basis. On the other hand, the Court argued, the Chapter 13 commitments do display a specific link with trade between the EU and Singapore. The Singapore Agreement seeks to ensure that the implementation of both sets of commitments (the trade commitments in the Singapore Agreement and the commitments derived from the multilateral conventions) are mutually compatible: it commits the parties, first, not to encourage trade by reducing the levels of social and environmental protection below the standards laid down by those international commitments, and second, not to apply those standards in a protectionist manner.

Further, a material breach of the sustainable development commitments could lead to a suspension of the trade liberalisation provisions.Footnote ⁶³ This somewhat surprising conclusionFootnote ⁶⁴ turns chapter 13 of the agreement into a form of sustainable development conditionality, potentially strengthening its effect:

Since the need to strengthen the sustainable development dimension of trade agreements has been a source of criticism of EU trade policy this development may be welcomed.Footnote ⁶⁶ However, as Ankersmit has pointed out, given that the Singapore Agreement’s dispute settlement procedures do not apply to chapter 13,Footnote ⁶⁷ it is not clear how this possibility of suspension would work in practice.Footnote ⁶⁸ One may also point out that if the possibility of conditionality-based suspension is taken seriously then the Court’s contention that chapter 13 does not affect the scope of the obligations under the international agreements it refers to, and therefore does not impose new obligations on the parties, becomes harder to maintain.Footnote ⁶⁹

In defining the scope of the CCP and its relation to internal competences the Court has thus developed a distinction. The CCP may encompass clauses which refer to standards established in existing agreements binding the EU; such clauses, according to the Court, preserve the right of the parties to set their own standards in conformity with their international obligations, and in accordance with their own internal legislative procedures.Footnote ⁷⁰ Article 207 TFEU does not, on the other hand, provide a basis for the harmonisation of labour or environmental standards; in accordance with Article 207(6) TFEU such harmonisation would require a separate legal basis.Footnote ⁷¹

Exclusive external competence under Article 3(2) TFEU

Notwithstanding the breadth of the CCP, two substantive sets of commitments in the Singapore Agreement fall outside its scope: transport services and non-direct foreign investment. In these fields, external powers must be implied from internal policy competences on the basis of Article 216(1) TFEU, and their exclusivity determined on the basis of Article 3(2) TFEU, provisions which represent an attempt to codify pre-existing case law on the existence and exclusivity of implied external competences. The basis for implied external competence is essentially two-fold.Footnote ⁷² First, it may be based on the existence of Union acts (‘common rules’) which may be affected if the member states act externally independently of the EU; it is thus based on the pre-emptive rationale that derives from the AETR case.Footnote ⁷³ Second, it may be based on an effectiveness (effet utile) rationale, where an external agreement is necessary to achieve the objectives established in the Treaties. Over the years an accretion of case law has built up over these two bases for external competence, and in particular over the conditions under which common rules may be ‘affected’ (what is sometime referred to an the ‘AETR-effect’). And the drafters of the Lisbon Treaty reflected the tendency of early case law to elide the distinction between the existence of implied external competence and its exclusivity by including in the Treaties two provisions, very similar but not identical in wording, one on the conditions under which the EU possesses the competence to conclude an external agreement (Article 216(1) TFEU), and one on the conditions under which the EU possesses an exclusive competence to conclude external agreements (Article 3(2) TFEU), with no indication as to their relationship. Thus since the Lisbon Treaty, two questions in particular have arisen: to what extent should these provisions be interpreted in line with earlier, pre-Lisbon case law? And how should we read the relationship between them? Opinion 2/15 helps to settle the answers to these questions – albeit with some ambiguity as regards the second, which has required a subsequent clarification. In the Opinion the Court did not take the apparently logical route of determining first the existence of competence and then deciding whether that competence is exclusive.Footnote ⁷⁴ Both the formulation of the Commission’s question and the uneasy relationship between the two provisions led the Court to start with the issue of exclusive competence, and only where exclusivity cannot be demonstrated did it, in some instances, go on to consider the possibility of nevertheless basing a shared competence on Article 216(1) TFEU.

Article 3(2) TFEU provides three possible conditions under which external competence may be exclusive (in addition to the a priori exclusive competences listed in Article 3(1) TFEU): when the conclusion of the international agreement is provided for in a legislative act of the Union; when it is necessary to enable the Union to exercise its internal competence; and in so far as its conclusion may affect common rules or alter their scope. These are based on the two rationales already mentioned: the first and third are linked to the existence of internal EU legislation or common rules (the ‘AETR-effect’); the second is the effet utile rationale. In Opinion 2/15, in line with its earlier case law, the Court first confirmed that these conditions reflect the pre-Lisbon case law,Footnote ⁷⁵ and that Article 3(2) TFEU should therefore be read in the light of that case law, as well as the cases on Article 3(2) TFEU that post-date the Lisbon Treaty.Footnote ⁷⁶ The focus in the Opinion is on the final limb of Article 3(2) TFEU – whether the provisions of the Singapore Agreement ‘may affect common rules or alter their scope’. Without summarising each step of the Opinion, the following points may be noted.

First, the Court follows its own injunction to engage in ‘a comprehensive and detailed analysis’Footnote ⁷⁷ of the different provisions of the agreement and the relevant EU legislation which may be affected. It looks at each type of transport in turn and its reasons for finding exclusive competence in each case are not identical.

Second, the EU’s ‘common rules’ may be affected where the agreement grants rights to EU operators under different conditions from those derived from existing EU legislation. Here, in the context of maritime transport, the Court pointed out that EU nationals established in Singapore would possess maritime transport rights as a result of the Singapore Agreement without necessarily operating under an EU Member State flag; the EU regulation in question would thereby be affected.Footnote ⁷⁸

Third, where an agreement provides for the application between the parties of rules that ‘overlap to a large extent with the common EU rules applicable to intra-Community situations’, whether or not there is any contradiction between them, that agreement may affect or alter the scope of the common rules.Footnote ⁷⁹ Thus provisions which essentially extend the benefits of internal market legislation to a third country are likely to satisfy the conditions for exclusive competence.

Fourth, the provisions relating to inland waterways transport services are of such limited scope that they may be disregarded without enquiring whether or not the conditions of Article 3(2) TFEU are fulfilled.Footnote ⁸⁰ The Court here applies in the context of exclusivity an approach, based on the distinction between predominant and incidental or ancillary purposes of an act, developed in the context of determination of legal basis.Footnote ⁸¹ Provisions which are regarded as purely ancillary or of very limited scope may not require a separate legal basis; it seems logical therefore that they should not by themselves alter the competence balance of the agreement.

Finally, in considering the possible exclusivity of competence relating to non-direct foreign investment, the Court follows the Advocate General in rejecting the Commission’s argument that the ‘common rules’ which may be affected under Article 3(2) TFEU may include rules laid down in primary EU law. Unusually (it is the only Treaty freedom to do so) Article 63 TFEU abolishes restrictions on the movement of capital between the EU and third countries as well as between EU member states,Footnote ⁸² and the Commission argued that this is a ‘common rule’ which would be affected by unilateral Member State action on investment. The Court put forward two reasons for rejecting the Commission’s position. The first is the case law history, what the Court refers to as ‘the reasoning inherent in the rule’ of the final phrase of Article 3(2) TFEU: in the AETR case, from which this rule ultimately derives, the common rules referred to were found in secondary legislation; exclusive external competence was based on the legislative exercise (and not only the existence) of an internal competence.Footnote ⁸³ While it has certainly always been assumed that the ‘common rules’ referred to are to be found in secondary legislation, this reasoning is not perhaps in itself completely convincing given the nature of the obligation in Article 63 TFEU. An alternative more convincing reason is suggested by the Advocate General: to derive exclusive competence directly from Article 63 would in effect be to hold that Article 63 TFEU creates an a priori exclusive competence (one that does not depend on EU action); a priori exclusivity is based on Article 3(1) not Article 3(2), and the list of such competences is closed.Footnote ⁸⁴ The Court’s second reason is, in my view, legally stronger: an international agreement cannot ‘affect’ or ‘alter the scope’ of a Treaty rule (such as Article 63 TFEU) since according to the hierarchy of norms in EU law, primary Treaty law takes precedence over international agreements.Footnote ⁸⁵

Neither the Advocate General nor the Court considered the possible application of the second limb of Article 3(2) TFEU, on the ground that the Commission did not argue the point. According to this test, competence to conclude an agreement will be exclusive when its conclusion ‘is necessary to enable the Union to exercise its internal competence’.Footnote ⁸⁶ It might indeed have been argued that in order to give full effect to the Treaty prohibition of restrictions on the movement of capital between the EU and third countries, the agreement of the third country – and thus an agreement with the third country – is necessary. The Court’s explanation of this ground of exclusivity in Opinion 1/03 could well be applied to the movement of capital:

It is well known, however, that although this ground of exclusivity continues to be mentioned by the Court, and is now enshrined in the Treaties, it has not been successfully pleaded since the ruling in Opinion 1/76.Footnote ⁸⁸ Tellingly, in the present case the Court went on (again following the Advocate General) to use the similarly-worded provision in Article 216(1) TFEU,Footnote ⁸⁹ together with Article 63 TFEU, as a basis for concluding that the EU possesses a shared external competence in relation to non-direct foreign investment.Footnote ⁹⁰ Apart from raising the question why the condition in Article 216(1) is satisfied while the (similar) condition in Article 3(2) is not, this step by the Court is interesting in two ways. First, while Article 216(1) TFEU codifies the pre-Lisbon case law on implied external powers, the tendency has been, in institutional practice and the Court’s judgments, to omit any reference to this provision as part of the legal basis for external agreements.Footnote ⁹¹ Here, an explicit reference is made to Article 216(1) but there is no suggestion that it should also be included as a legal basis.Footnote ⁹² Second, the Court concluded that competence over non-direct foreign investment is shared, not exclusive. What are the implications of this finding for the central question asked by the Commission: does the Union have the requisite competence to sign and conclude alone the free trade agreement with Singapore?

Does shared competence necessarily imply mixity?

As we have just seen, the part of chapter 9 of the Singapore Agreement which deals with non-direct foreign investment falls within shared EU competence. The Court concluded this part of the Opinion with this surprising statement: ‘It follows that Section A of Chapter 9 of the envisaged agreement cannot be approved by the European Union alone.’Footnote ⁹³ As a number of commentators have pointed out, this seems to imply that shared external competence necessitates mixity.Footnote ⁹⁴ This would indeed be a novel outcome, overturning many years of practice, and highly significant for future EU treaty-making. Although some types of agreement are routinely concluded as mixed agreements (in particular important Association agreements), many agreements in fields of shared competence are regularly concluded by the EU alone – agreements in the fields of the Common Foreign and Security Policy, for example, as well as justice and home affairs – and this practice has never been questioned.

Certainly there may be cases (although they are likely to be rare) where competence is shared in the sense that there are parts of the agreement which fall outside EU competence altogether; here, mixity would be required.Footnote ⁹⁵ But where the EU possesses competence over the whole agreement albeit shared (in whole or part) with the member states, then mixity is a matter of political choice and will depend on many factors, including the political importance of the agreement. This was in fact the position assumed by the Council in this case, as also by the Advocate General.Footnote ⁹⁶ It seems, to say the least, unlikely that the Court would have decided to take such a decisive step without any explanation or rationale. Despite the unfortunate choice of words, the Court must have meant that Member State participation could not be excluded as a matter of law, in other words that the Singapore Agreement could not be concluded by the EU alone unless the member states so decide, the latter being a political decision it could not comment on. Indeed, the member states submitting comments all argued that the EU should not conclude the Singapore Agreement alone and this was also clear from discussions in Coreper.Footnote ⁹⁷ The Court therefore knew that in practical terms a finding of shared competence meant that the agreement would be mixed, and this presumably influenced its choice of words. The conclusion of the Advocate General is also more nuanced: ‘Since not all of those parts fall within the scope of the European Union’s exclusive external competences, the EUSFTA cannot, on the basis of the European Union’s exclusive competences, be concluded without the participation of the Member States.’Footnote ⁹⁸ It should also be noted that although the Court uses the same form of words more than once in the Opinion,Footnote ⁹⁹ its operative conclusions refer simply to shared competence with no indication as to who should or could conclude the agreement.Footnote ¹⁰⁰

This reading of the troublesome phrase has now been confirmed by the Court itself, in an explanatory paragraph in a subsequent case.Footnote ¹⁰¹ Dismissing an argument that Articles 3(2) and 216(1) TFEU are effectively coterminous (that is, that the EU only has an implied external competence where the one of the conditions set out in Article 3(2) TFEU is satisfied) the Court held:

Although the Court here refers specifically to competence over non-direct foreign investment, its comment clearly applies to all the cases of shared competence in the Singapore Agreement;Footnote ¹⁰³ we will return to this point below in the context of ISDS. For now it is enough to conclude that the short-lived confusion over whether a finding of shared competence necessarily requires mixity, generated by the ambiguous wording in the Opinion, has been settled in favour of the status quo ante: in cases where the EU possesses competence over the whole of an agreement, albeit partly or wholly shared competence, mixity is not a legal requirement; it is a choice which will ultimately be made by the Council in determining the negotiating directives for the agreement.

The member states’ bilateral investment treaties

The Singapore Agreement is intended to replace a number of bilateral investment treaties that some Member States have concluded with Singapore. The agreement takes the unusual step of providing that Member State-Singapore bilateral investment treaties will cease to have effect on the entry into force of the Singapore Agreement and that all rights and obligations under these bilateral investment treaties will be ‘replaced and superseded’ by the free trade agreement.Footnote ¹⁰⁴ The Advocate General and the Court take very different views of the allocation of competence for this provision, adopting different frames of analysis; in the Advocate General’s view, this clause lay within the Member States’ exclusive competence (and thus required their participation in the agreement); in the Court’s view the clause lay predominantly within exclusive EU competence.

The Advocate General’s starting point is that responsibility for ensuring compatibility between a Member State’s obligations under EU law and under its agreements with third countries is a matter for that Member State, on the basis of the primacy of EU law and the duty of sincere cooperation under Article 4(3) TEU, and in line with international treaty law. In the case of agreements entered into prior to accession the EU, which applies to some of the bilateral investment treaties in question, Article 351(2) TFEU likewise puts the responsibility on the Member State to eliminate any incompatibilities; it assumes that the Member State will remain party to the third country agreement with responsibility to bring either the incompatibility, or if necessary the agreement, to an end.Footnote ¹⁰⁵ According to the Advocate General this division of responsibility is not affected by a change in competence distribution (foreign direct investment having become exclusive EU competence). As she says,

The Court of Justice, however, decided that, by virtue of its now-exclusive competence over foreign direct investment as regards third countries, the EU had succeeded to the member states’ third country bilateral investment treaties. Relying on its case law derived from the International Fruit Company case,Footnote ¹⁰⁷ often referred to as the doctrine of functional succession, the Court concluded that ‘the European Union replaces the member states so far as concerns international commitments entered into in fields which, like that of foreign direct investment, fall within its exclusive competence’.Footnote ¹⁰⁸ It thus, in contrast to the Advocate General, bases itself entirely on EU law.

It will be noted that the Court is careful to limit this ruling to foreign direct investment; as we have seen, non-direct foreign investment falls within shared competence. The Court alludes to this distinction but does not regard it as a barrier to the application of functional succession.Footnote ¹⁰⁹ Overall, although the outcome in this particular case is unexceptional, the Opinion’s position on this issue is, in my view, problematic. The argument that termination of the third country bilateral investment treaties falls within exclusive Member State competence is not wholly convincing,Footnote ¹¹⁰ but there are good reasons why functional succession has been so sparingly applied, and the Court here extends it in two ways: by applying it to Member State agreements which fall partly within shared competence,Footnote ¹¹¹ and by applying it in the context of revoking rather than implementing those agreements. The doctrine of functional succession is not intended to alter the international law position of third parties, but rather to express an EU law-derived obligation on the Union to respect an agreement to which it is not formally a party. The Advocate General refers in the passage quoted above to ‘the fundamental rule of consent in international law-making’;Footnote ¹¹² in the International Fruit Company case (in which the Court held that the then EEC was bound by the General Agreement on Tariffs and Trade, despite not at that time being a party to the agreement) the consent of the other parties to that Agreement, their acceptance of the EEC as an interlocutor within the Agreement, was an important step in its argument.Footnote ¹¹³ Here too the Court refers to Singapore’s consent to the clause in the Singapore Agreement, but not as a pre-condition for the operation of functional succession; instead it is given as one reason for rejecting the relevance of Article 351 TFEU: according to the Court, since Singapore has consented to the clauses in the Singapore Agreement its rights under the Member State bilateral investment treaties do not therefore need the protection which is given by Article 351 TFEU.Footnote ¹¹⁴ This is in fact a non sequitur, since Article 351(2) TFEU actually envisages that the member states will attempt to remove any incompatibility between EU law and a prior agreement through negotiation with that third country, if necessary assisted by the other member states and (according to the Court) by the Commission.Footnote ¹¹⁵ Indeed it is true that functional succession and Article 351 TFEU serve different purposes: whereas Article 351 TFEU allows the member states to fulfil obligations to third countries arising out of prior (pre-accession) treaty commitments, the doctrine of functional succession is an exceptional mechanism that allows the EU, with the consent of the third country party(ies), to fulfil Member State obligations under agreements which have moved into exclusive EU competence, and provides that the agreement should be regarded as binding on the EU as a matter of EU law. Both, however, are concerned with ensuring that obligations to third countries are respected, notwithstanding transfers of competence between the EU and its member states, and neither is intended to alter the existing state of obligations under international law. They can indeed be seen as complementary. Functional succession has never been regarded as a mechanism for bringing Member State obligations towards third countries to an end. The Court’s approach to the question here effectively deprives the member states of any power to manage their own existing international law obligations when a field moves into exclusive EU competence, even in cases where there is no envisaged EU external agreement;Footnote ¹¹⁶ this is the kind of situation which may create problems for the third country and which Article 351 TFEU was designed to avoid. In practice in this particular instance, since the Member State bilateral investment treaties fall partly within shared competence, and since the intention is to replace them with the Singapore Agreement, the application of functional succession will not have a dramatic effect; however the reasoning used here would necessarily apply also to Member State agreements which fall entirely within exclusive EU competence and would be triggered irrespective of the intention of the EU to conclude a replacement. An approach more reflective of legal and practical realities would have been to accept that it was for the member states to terminate these agreements,Footnote ¹¹⁷ but that there is nothing to prevent them arranging for the Union to do so on their behalf, through a clause in the Singapore Agreement.Footnote ¹¹⁸

Competence in relation to ISDS

A final issue of competence deserves a brief comment. In addition to a relatively complex but largely traditional institutional and inter-party dispute settlement structure, the Singapore Agreement contains provision for investor-state dispute settlement. Unlike the EU-Canada Comprehensive Economic and Trade Agreement, which adopts the EU’s newer investment court-based form of ISDS,Footnote ¹¹⁹ the ISDS scheme in the Singapore Agreement is based on arbitration. Generally speaking, institutional and dispute settlement provisions designed to ensure the effective implementation of substantive obligations will be regarded as ancillary, and will therefore ‘follow’ the competence distribution of the substantive provisions and share their legal basis.Footnote ¹²⁰ Although the Advocate General applied this reasoning also to the ISDS scheme,Footnote ¹²¹ the Court did not. Under the terms of the Singapore Agreement the member states (and indeed the EU) are deemed to consent to investor-state arbitration, and as a result a Member State would be unable to oppose the decision of an investor-claimant to submit a dispute to arbitration. According to the Court,

The Court goes on to hold that competence in relation to ISDS in the Singapore Agreement is not exclusive to the EU, but shared.Footnote ¹²³ This does not differ in result from a finding that ISDS is ancillary to the substantive obligations on investment, since as we have already seen competence as regards non-direct investment protection is also shared. But one result of the Court’s refusal to treat the ISDS provisions as ancillary to the substantive obligations is that they then need a legal basis in their own right, and the Court surprisingly makes no mention of what this legal basis might be.Footnote ¹²⁴ Presumably the legal basis would reflect that of the substantive provisions (that is, Article 207 TFEU for direct investment and Article 63 TFEU for non-direct investment), but the omission is odd given that allocation of competence normally follows legal basis.

In the passage just quoted the Court says that the ‘consent’ of the member states is required. Might this signal that the member states need to be party to the agreement insofar as it contains provision for ISDS, on the ground that ISDS vis-à-vis the member states is a matter of (exclusive) member state competence, whereas ISDS vis-à-vis the EU is a matter of (exclusive) EU competence?Footnote ¹²⁵ This would result in shared competence over ISDS, but of a different type from that applying to non-direct investment.Footnote ¹²⁶ However, in the final statement of its Opinion, the Court does not distinguish between different types of shared competence and it is unlikely to have intended to do so here. Indeed, this is not one of the passages in the Opinion where the Court concludes that the agreement ‘cannot be approved by the EU alone’. Shared competence, then, does not necessarily require member state participation; the Member States may consent to the agreement being concluded by the EU alone, if it is their political choice not to insist on a mixed agreement. As mentioned above, the ‘clarification’ which followed Opinion 2/15 over the need for a mixed agreement in all cases of shared competence, should apply also to shared competence over ISDS.Footnote ¹²⁷