The Opinion procedure under Article 218(11) TFEU: Reflections in the light of Opinion 1/17

This paper explores aspects of Opinion 1/17, handed down by the Court of Justice on 30 April 2019, as an example of the specific procedure of Article 218(11) TFEU and the Court’s evolving practice. The prior Opinion procedure serves to prevent the complications that would arise, both internally and externally, if an international treaty were concluded by the EU and then subsequently found to be incompatible with primary EU law. Opinion 1/17 raised issues of institutional and substantive compatibility, in the form of the principle of autonomy and compliance with the Charter of Fundamental Rights. In handling both, the Court’s approach was characterised by its insistence on the reciprocal nature of the CETA relationship, the separation of the CETA system from EU law, the ability of the EU to engage with such independent and reciprocal dispute settlement processes, and its assessment of the way in which the agreement would be implemented. The Opinion represents an example of the explanatory or didactic form of reasoning typical of recent Opinions, as well as breaking new ground in its conception of an ‘envisaged’ agreement.


Introduction
Opinion 1/17, handed down by the Court of Justice on 30 April 2019, 1 concerned the compatibility with the EU Treaties of the projected Comprehensive Economic and Trade Agreement between the EU and Canada (CETA). 2 The Opinion had been requested by Belgium under Article 218(11) Treaty on the Functioning of the European Union (TFEU), the only judicial procedure specifically directed at international agreements, which is -uniquely -embedded in the procedural rules for treaty-making, rather than the provisions on the jurisdiction of the Court. 3 The purpose of this 'exceptional procedure' of ex ante constitutional review of the EU's international agreements, was explained by the Court on the first occasion the procedure was used, 4 and this rationale was then used by the Court to justify its interpretation of the scope of the questions that could be raised. The same phraseology is still used by the Court in determining its tasks under the Opinion procedure, 5 and is worth quoting in full: It is the purpose of [Article 218(11) TFEU] to forestall complications which would result from legal disputes concerning the compatibility with the Treaty of international agreements binding upon the Community. In fact, a possible decision of the Court to the effect that such an agreement is, either by reason of its content or of the procedure adopted for its conclusion, incompatible with the provisions of the Treaty could not fail to provoke, not only in a Community context but also in that of international relations, serious difficulties and might give rise to adverse consequences for all interested parties, including third countries. For the purpose of avoiding such complications the Treaty had recourse to the exceptional procedure of a prior reference to the Court of Justice for the purpose of elucidating, before the conclusion of the agreement, whether the latter is compatible with the Treaty. 6 The procedure, according to this statement, is not primarily about protecting rights or institutional prerogatives, 7 but serves to prevent the 'complications' that would arise, both internally and externally, if an international treaty were concluded by the EU and then subsequently found to be incompatible with primary EU law (the EU Treaties). Internally, since an agreement concluded by the EU is a source of EU law which takes precedence over secondary legislation, 8 and the Union would thus be faced with a conflict in its hierarchy of norms. Externally, since while the agreement would still be binding on the EU as a matter of international law, a finding of incompatibility would at the least make it difficult for the EU to fulfil its obligations and would most likely entail the EU's withdrawal from (and perhaps an attempt to renegotiate) the agreement. 9 1 Opinion 1/17, 30 April 2019, ECLI:EU:C:2019:341. 2 Council Decision 2017/37 of 28 October 2016 on the signing on behalf of the European Union of the Comprehensive Economic and Trade Agreement (CETA) between Canada, of the one part, and the European Union and its Member States, of the other part; OJ 2017 L 11/1. 3 Article 218 (11) TFEU provides: '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. Where the opinion of the Court is adverse, the agreement envisaged may not enter into force unless it is amended or the Treaties are revised.' The provision was originally included in Article 228 EEC and has only been slightly amended, to include the European Parliament among those entitled to request an Opinion, and to clarify the consequences of an adverse ruling. On the Opinion procedure generally, see further S Adam, La procédure d'avis devant la Cour de justice de l'Union européenne ( The Opinion procedure has existed since the original Treaty of Rome. 10 As we have just seen, the first Opinion was not requested for nearly 20 years, in 1975, and it is relatively rarely used compared with other judicial procedures. Only 22 Opinions have been delivered in total, and never more than three in one year. 11 However, its use is becoming a little more common: of the 22 existing Opinions, half date from 2000 and about a third have been delivered in the last decade. Despite their small number, Opinions have proved important in developing not only the fundamental principles of EU external relations law, including the principle of autonomy and its application in particular to international dispute settlement, 12 but also the general constitutional law of the EU. For example, the characterisation of the EEC Treaty as 'the constitutional charter of a Community based on the rule of law'; 13 the role of national courts in the Union legal order and their relationship to the Court of Justice; 14 and the identification of the 'specific characteristics' of the EU's 'constitutional structure'. 15 Opinion 1/17 takes its place in this tradition. The structure and style of its reasoning follows that of earlier Opinions, especially in its explanatory and almost didactic character; it attempts to systematise and explain existing case law on the principle of autonomy in the context of international dispute settlement; it builds upon Opinion 1/15 in offering an assessment of the compatibility of an international agreement with the Charter of Fundamental Rights; and it offers an innovative approach to compatibility which draws upon the concept of an 'envisaged' agreement. The following brief assessment is certainly not intended to offer a complete analysis of the Opinion, different aspects of which are explored in other contributions to this Special Issue; it rather seeks to draw attention to a few aspects of the Opinion as an Opinion, as an example of the specific procedure of Article 218(11) TFEU and the Court's evolving practice.

The context and timing of Opinion 1/17
The request for an Opinion was made by Belgium on 7 September 2017. The request concerned the compatibility with the Treaties, 'including with fundamental rights', of provisions of the CETA dealing with the resolution of investment disputes between investors and states (ISDS). Requests from Member States are relatively unusual, but, interestingly, of the four that have been made by Member States three have come from Belgium. 16 In many cases, even where the request comes from a Member State, the genesis of the request lies in different views over questions of competence or the institutional implications of the envisaged agreement. It is rare for this procedure to be used to raise issues of substantive policy choice, 17 and this is the first time where the impetus for the request has come, essentially, from domestic debate within a Member State. The one case where Germany made the request arose out of a difference of view within the Council, as well as with the Commission, over the substantive content of the agreement and the policy behind it. 18 In this case, however, the request from Belgium arose out of the concerns within the Member State itself (notably within one of its regional parliaments) over the substance of the agreement. 15 Opinion 2/13 (n 12) paras 165-8. 16 Ruling 1/78 (EAEC) (n 11) and Opinions 2/92 (OECD National Treatment Instrument), ECLI:EU:C:1995:83, and 1/17 (CETA) (n 1) were requested by Belgium; Opinion 3/94 (Framework Agreement on Bananas) (n 7), was requested by Germany. The Commission has initiated the procedure most frequently (14 times), the Council three times and the European Parliament once. 17 Opinion 1/15 (n 5) on the PNR agreement with Canada, requested by the European Parliament, might be said to be another example. 18 Opinion 3/94 (n 7) in which Germany questioned the compatibility with the Treaty of the Framework Agreement on bananas; the request was deemed inadmissible as the agreement had already been concluded and the Court therefore did not discuss the substantive issues raised.
In this case, as in a few others, 19 the request was made after the signature of the agreement but before its formal conclusion. According to Article 218(11) TFEU, the agreement must be 'envisaged', that is, not yet formally concluded by the Union. 20 There is at least a potential risk, once a request has been made, of an agreement being concluded before the Court has delivered its Opinion, thereby rendering the agreement no longer 'envisaged', and in fact, on two occasions this has happened. In the first, the Framework Agreement on bananas had been signed before the request for an Opinion was made by Germany, and was then concluded as part of the Uruguay Round package of World Trade Organization (WTO) agreements before the Court had delivered its Opinion. The Court, relying on the rationale of the Opinion procedure as well as on the wording of the provision, 21 declared the request inadmissible as 'devoid of purpose'. 22 On the second occasion, the European Parliament requested an Opinion on the first Passenger Name Record (PNR) agreement with the US, but the Council concluded the agreement before the Court had delivered its Opinion. The Parliament then withdrew its request and subsequently successfully challenged the Council decision concluding the agreement. 23 As the CETA is a bilateral mixed agreement, there was no real risk that the Council would conclude the agreement before the Court had given its Opinion. 24 Nevertheless, at the time when signature was agreed Belgium made a political commitment that it would not ratify CETA before the Court had given its Opinion, and it also ensured that the provisions on ISDS about which it sought the Court's Opinion were not included in the scope of provisional application of CETA pending its conclusion. 25 In a Statement in the Council minutes, Belgium said both that it intended to ask for an Opinion and that it might not be able to ratify the CETA, depending on the ratification process in its regional assemblies. 26 The CETA Opinion is therefore notable for the fact that it was made clear by the Member State asking for the Opinion that its own ratification of the agreement, and not only its conclusion by the Union, was linked to the result of the Opinion procedure.

The subject-matter and structure of the Opinion
While the basic structure of the Opinion was established rather early and has changed little, the Opinions delivered since 1975 fall into four broad chronological groups, which reveal an evolution in approach and style. 27 In the first group of Opinions handed down in the 1970s 28 the Court established the basic parameters of the procedure and its potential as a forum for resolving all constitutional questions around treaty-making (and resisting claims that these issues should simply be left to political agreement). There is an emphasis on ensuring the development and exercise of the international identity of the EU, separate from its Member States (the inception of the principle of autonomy). Then, during the 1990s a second group of Opinions were handed down which establish key concepts of EU constitutional, as well 19 See e.g. Opinion 1/94 (WTO), ECLI:EU:C:1994:384; Opinion 1/08 (GATS schedules), ECLI:EU:C:2009:739; Opinion 1/15 (PNR Canada) (n 5). 20 Opinion 1/94 (n 19) para 12. 21 See above at note 6. 22 Opinion 3/94 (n 7). The Court pointed to the possibility of a subsequent challenge to the legality of the concluding decision: see Opinion 3/94 at para 22; see further Case C-122/95 Germany v Council, ECLI:EU:C:1998:94. 23 Opinion 1/04, withdrawn. See above at note 9. 24 Bilateral mixed agreements will not generally be concluded by the Council until all Member States have completed their ratification processes. In the case of CETA, unusually, a Statement by the Council was attached to the minutes of the Council decision on signature of the agreement, alluding to the possibility that the ratification of CETA may fail 'permanently and definitively' as a result of domestic constitutional processes, and providing that in such case provisional application will be terminated. as external relations, law. 29 In these Opinions the Court started its practice of beginning with a summary of the relevant field of law or the fundamental characteristics of the Community legal order, a practice which has become an important feature of these non-contentious cases. The style of these Opinions is more overtly designed not only to answer the specific question, but also to build a jurisprudence constante, the Opinion developing into one of the instruments used by the Court to establish itself as a constitutional court. The third group of Opinions, delivered in the first decade of the 2000s, 30 follow the style developed in the 1990s, consolidating established lines of case law, and providing statements or summaries which are often cited in later cases and used as a starting point for further developments. 31 Opinion 1/17 is the most recent in the fourth group of Opinions, delivered since the coming into force of the Lisbon Treaty. 32 In these and other cases decided in the last decade, the Court has explored the impact of the Lisbon Treaty on EU external relations, developing its legal assessment of the altered constitutional framework for external action. 33 One feature of the more recent Opinions is the publication of a single Advocate General's opinion (or 'view'), replacing the closed hearing of all the Advocates General that was the practice until the revision of the Court's rules of procedure in 2012. 34 The Advocate General's opinions, as in other procedures, provide an opportunity for alternative perspectives, and the potential for deepening the analysis. 35 The increasing use of the Opinion procedure is demonstrated by the fact that seven Opinions have been delivered over the last nine years, almost a third of the overall total. Of these, four dealt with competence and exclusivity-related issues, and four with issues of compatibility, a distribution reflecting the practice in earlier Opinions. Opinions are also becoming longer: Opinion 1/17 is 245 paragraphs; Opinion 2/13 was 258 paragraphs and Opinion 2/15 was 305 paragraphs; in contrast, Opinion 1/94 was 110 paragraphs and Opinion 1/76 was all of 22 paragraphs long. Sometimes the length is reflective of the complexity of the agreement under scrutiny, but the longer Opinions also reflect the space given to summarising the submissions and their didactic and expanatory character, giving space to a general exposition of the law.
The Court's jurisdiction under Article 218(11) TFEU relates to the compatibility of the envisaged agreement with the EU Treaties. Since the very first Opinion, the Court has read this as extending to 'all questions capable of submission for judicial consideration' that may raise doubts as to the validity of the agreement as a matter of Union law. 36  types of issue have been raised in Opinion procedures: competence questions, 38 and those relating to other compatibility issues. 39 Until recently, the latter essentially concerned institutional compatibility, such as decision-making or judicial autonomy, rather than compatibility with substantive EU law. Opinion 3/94 (Germany's request on the bananas agreement) raised a substantive compatibility issue (discrimination) but the Court rejected the request on grounds of admissibility and did not get to the merits. Opinion 1/12 (on the Anti-Counterfeiting Trade Agreement, ACTA) also raised substantive compatibility issues (fundamental rights), but was withdrawn when the proposed ACTA failed to get the consent of the European Parliament. The first Opinion actually to address substantive compatibility was Opinion 1/15, on the PNR agreement with Canada, which discussed Articles 7 and 8 of the Charter of Fundamental Rights. Opinion 1/17 raised both institutional and substantive aspects of compatibility: the former in the form of the principle of autonomy, and the latter primarily in terms of compliance with the Charter of Fundamental Rights. We will consider its reasoning on each of these further in the next section. As far as substantive compatibility is concerned, it is difficult to draw conclusions on the basis of only two Opinions, but its recent use suggests that it is likely to become a more important aspect of this form of prior constitutional review, and it is perhaps significant that in each case the compatibility issues were linked to Charter rights.
The basic structure of an Opinion under what is now Article 218(11) TFEU was settled as early as Opinion 1/76, although the single consecutively numbered text dates from Opinion 1/03. It comprises the statement of the request for an Opinion; a summary and analysis of the envisaged agreement and sometimes of its legal context; a summary of the appraisal of the requesting institution or Member State and of the main observations submitted to the Court; and finally the position of the Court. Opinion 1/17 follows this structure: first, it cites and summarises the relevant provisions of CETA, together with the parties' Joint Interpretative Statement; then it gives a summary of the Belgian concerns, followed by a summary of observations. The Court's position 40 is divided into the same three issue areas as the Belgian position and the Advocate General's opinion, that is (i) the compatibility of the envisaged ISDS mechanism with the autonomy of the EU legal order; (ii) the principle of equal treatment and the requirement of effectiveness; (iii) the right of access to an independent and impartial tribunal.
Opinion 1/17 adopts the practice of starting with a general disquisition on the relevant legal issues. This has a quasi-didactic function, summarising the fundamental principles in a general way in a passage capable of being cited in subsequent cases. Opinion 2/13 is a very clear example 41 but this approach can be identified from the first European Economic Area (EEA) Opinion onwards. 42 In Opinion 1/17, this general summary of the law comes in three stages: each of the three parts of the argument, before discussing the specific agreement, starts with a section on 'principles' which summarises earlier case law and highlights the points the Court wants to emphasise. Thus, the 'principles' section at the start of the section on autonomy uses the language of Opinion 2/13, as repeated since -the Court cites Wightmanon autonomy as protective of the EU's 'constitutional framework', and summarising the components of this framework. 43

Content and reasoning of the Opinion
In what follows we will highlight just two features of the Court's reasoning in Opinion 1/17: its approach to the principle of autonomy, and the innovative elements in its handling of the issue of compatibility. 38 Competence questions have been raised in 15 cases, in only one of which did the Court hold that the Community/Union had no competence at all (Opinion 2/94 on the ECHR). 39 Compatibility has been raised in nine cases: in five cases the agreement was found to be incompatible (Opinions 1/76, 1/91, 1/09, 2/13 and 1/15), in three cases compatible (Opinions 1/92, 1/00, 1/17) and in one case inadmissible (Opinion 3/94). 40

The principle of autonomy
The autonomy of the Community (and now Union) legal order has emerged as a structural principle of EU external relations over a series of cases, mainly though not exclusively concerned with external dispute settlement procedures. 44 As Lenaerts has pointed out, the principle of autonomy reflects both the claim, going back to Van Gend en Loos, that EU law is not simply international law, and the idea that EU law 'operates as a self-contained system of norms'. 45 Opinion 1/17 represents an attempt by the Court to explain more systematically the implications of the autonomy principle, building upon earlier Opinions, from Opinion 1/76 to Opinions 1/91, 1/00, 1/09 and of course Opinion 2/13, as well as other cases such as Achmea. 46 It also develops earlier reasoning on decision-making autonomy into a substantive assessment of regulatory autonomy. 47 Other contributions discuss this aspect of Opinion 1/17 in more detail. 48 Here, I wish to draw attention to two specific features of the Court's reasoning.
First is the attention paid, more overtly than is usual, to the EU's position as an international actor and its ability to enter into reciprocal commitments with external partners. Advocate General Bot, while insisting that he will 'set aside, despite their significance' the political and economic aspects of the question before the Court, 49 nonetheless emphasised the policy context: the part played by the EU in promoting the reform of ISDS internationally, and the 'experimental and dynamic nature' of the ISDS model adopted for CETA. 50 The Court starts from the position that it understands 'the need to maintain the powers of the Union in international relations' and the consequent importance, in an arms-length bilateral relationship, of an independent and fully reciprocal dispute settlement system. 51 Second is the importance given to the independence of the ISDS system from the courts or tribunals of the parties, and the link forged between the independence of that system, the explicit denial of direct effect of the CETA, and reciprocity. Reciprocity has had a complex relation to the direct effect of the EU's international agreements. In Kupferberg the Court was not dissuaded from finding a clause in a free trade agreement to be directly effective by the possibility that the other party might not give it such effect. Such a potential imbalance would not, in its view, constitute a lack of reciprocity in implementation given the overall obligation to perform the agreement in good faith and the other institutional procedures established by the agreement. 52 In contrast, in Portugal v Council the Court used reciprocity in reasoning that the WTO does not have direct effect. The reciprocal nature of the obligations in the WTO agreements demands, in the Court's view, reciprocal dispute settlement and enforcement procedures. 53 In the case of CETA, the express exclusion of direct effect 54 is linked by the Advocate General (and impliedly by the Court) to the establishment of a reciprocal system of dispute settlement, independent of the domestic legal systems of the parties. Indeed the Court puts particular emphasis on the fact that the CETA's ISDS mechanism 'stands outside the EU judicial system'. 55 This becomes a factor in the Court's determination that the autonomy of that system is not threatened.
The Court builds on the Advocate General's analysis of reciprocity in the CETA, 56 developing its ideas on the relation between autonomy and dispute settlement within international agreements. The reciprocal nature of the agreement and the need to 'maintain the powers' of the EU with respect to its international partners underpin the choice of an ISDS mechanism which 'stands outside' both EU and Canadian domestic law. The relationship between the parties, in this external context, is characterised by reciprocity, as opposed to the relationship of mutual trust operating as a fundamental characteristic of the relationship between EU Member States. 57 The absence of direct effect reinforces the separation since it removes the option of bringing a complaint before a domestic court, 58 and since ISDS is only available to foreign investors there is no alternative mechanism allowing recourse to CETA in an intra-EU context. 59 The Court clearly regards such a complete separation as being less problematic than the attempts at integration between systems that were a feature of the EEA, the Patent Court and European Convention on Human Rights (ECHR) cases. 60 As long as the ISDS tribunal is not given the jurisdiction to interpret or apply EU law other than the CETA itself, nor powers to issue awards which would affect the operation of the EU institutions, the principle of autonomy will not be jeopardised. 61 So, having in earlier cases emphasised the threats to the autonomy of the EU legal order, the Court now makes it clear that the principle of autonomy need not hinder the EU from engaging with other international actors and protecting its position via international dispute settlement. It distinguishes, more clearly than in earlier cases, between intra-EU and extra-EU relations (governed by EU law and international law respectively), stressing the mutual trust operating in the former, as well as its democratically decided policy choices. Unlike the ECHR accession agreement, which left open the possibility of the Convention applying in intra-EU relations, the CETA regime cannot be relied on by EU investors within the EU: they cannot use the ISDS as they are not foreign investors in the EU, nor can they rely on CETA in EU domestic courts, given its lack of direct effect. 62 Both these features of the Court's reasoning on the principle of autonomy display an attention to the international identity of the EU; they are outward rather than inward facing. They are not defensive in nature; rather, the Court positions the EU as a (non-State) subject of international law, for whom international law is a separate system in which it participates, effectively on the same (reciprocal) level as its third State partners. However, other parts of its reasoning are more concerned with preserving the EU's (internal) constitutional framework, both institutional and substantive; it is to this aspect of the Opinion that we will now turn.

Assessing compatibility: The 'envisaged agreement'
As we have already seen, Opinion 1/17 addresses both institutional and substantive compatibility issues. In both cases, the Court adopts an innovative approach to assessing compatibility, basing its 54 Article 30.6.1 CETA provides that 'nothing in this Agreement shall be construed as . . . permitting this Agreement to be directly invoked in the domestic legal systems of the Parties'. 55 Opinion 1/17 (n 1) para 113. 56 Opinion 1/17, opinion of AG Bot (n 49) paras 77-85. 57 Opinion 1/17 (n 1) paras 128-9; Opinion 1/17, opinion of AG Bot (n 49) para 81. See further Contartese (n 47) p 7. 58 Opinion 1/17 (n 1) para 198. 59 ibid, para 181. 60 Opinion 1/92 (n 12); Opinion 1/09 (n 12); and Opinion 2/13 (n 12). 61 Opinion 1/17 (n 1) para 119. 62 ibid, para 181. The Court does not see this as problematic from the point of view of equal treatment as it argues that the situation of Canadian investors and EU Member State investors investing in the EU are 'not comparable': Opinion 1/17 (n 1) para 180. The argument is somewhat circular (the difference which may be regarded as contrary to equal treatment is itself used to support the case that the two are not comparable), but its use by the Court emphasises the separation between the CETA regime (applicable to EU Member State investors in Canada, but not within the EU) and EU law. judgment on a reading or understanding of how the CETA will be interpreted and applied (including by an ISDS tribunal), and on assurances taken from the Joint Interpretative Statement as to how the parties envisage its application and its limits. In doing so it effectively signals what, in its view, would be the correct interpretation or application of the agreement by the tribunal, although of course -as the Court also recognises -that tribunal will interpret the CETA independently and having regard to the rules and principles of international law applicable between the parties, and not EU law. 63 In line with the principle of autonomy and the separation between the CETA ISDS mechanism and EU law, the Court accepted that its own jurisdiction would not 'take precedence' over the jurisdiction of the CETA tribunal. 64 Despite such statements, throughout the Opinion the Court signals the ways in which it expects the agreement to be interpreted and implemented so as to ensure compatibility. While somewhat unusual, this is unexceptional in legal terms where it relates to implementation by the EU institutions. Thus, the Court reminds the institutions that the Union's position in Joint Committee decisions adopting binding interpretations of the CETA, which would be governed by Article 218(9) TFEU, must comply with EU primary law, 'in particular the principles that are part of that law and that are set out or clarified in the present Opinion'. 65 The relevant CETA provision 'cannot be interpreted' as permitting the Union to consent to decisions on interpretation in the CETA Joint Committee that would conflict with EU primary law, including the Charter of Fundamental Rights. 66 Note that the Court is here itself interpreting CETA in such a way as to ensure its compliant implementation -including in matters of interpretation -by the EU.
It is legally more surprising that the Court applies the same approach, relying on a specific interpretation of the envisaged agreement, to external actors, in particular the independent CETA tribunal. Thus, in the context of the need to ensure the Union's regulatory autonomy, the Court addresses the concern that the EU might need to alter its legislation following assessment by a tribunal 'of the level of protection of a public interest established, in accordance with the EU constitutional framework, by the EU institutions'. In such a case, it held, the agreement would undermine 'the capacity of the Union to operate autonomously within its unique constitutional framework'. 67 However, after reviewing provisions protective of public policy and the right to regulate in the CETA as well as the Joint Interpretative Statement, the Court concluded that this risk would not materialise. The powers of a CETA tribunal, it decided, 'do not extend to permitting them to call into question the level of protection of public interest determined by the Union following a democratic process'. 68 The reasoning here is striking in several ways. Regulatory autonomy is defended as necessary to protect the EU's specific constitutional balance in decision-making and the resulting regulatory choices. 69 As Contartese points out, this approach defends the outcomes of the EU's internal legislative process against the (potentially different) balance struck in its international decision-making. 70 Second, it is assumed that the tribunal's view of the extent of public policy discretion, and its exercise in a specific context, will coincide with that of the Union itself. More importantly, the Court predicated its finding of compatibility on its own determination of the scope of (and limits to) the tribunal's jurisdiction, declaring that the CETA tribunal has 'no jurisdiction' to declare the level of protection of a public interest established by EU measures incompatible with the CETA. 71 Similarly, in considering whether the CETA may adversely affect the effectiveness and enforcement of EU competition law, the Court assumed that the CETA tribunal simply could not find a breach of CETA where EU competition law has been correctly applied. An award by a CETA tribunal, holding that an EU competition decision is in breach of CETA, is 'unimaginable' where the competition rules have been correctly applied by the Commission or the competition authority of a Member State. 72 Again we find an assumption that CETA and EU law, although subject to separated dispute settlement and enforcement, will not diverge.
When considering compatibility with the Charter of Fundamental Rights, the Court relied on assurances as to how CETA would be implemented by both the Joint Committee and the EU. In discussing the issue of access to ISDS under CETA for small and medium-sized enterprises (SMEs), the Court recognised that commitments to facilitate such access were not contained in the CETA in legally binding form. However, it relied on undertakings made by the Council and Commission in Statement No 36, which was, with others, entered into the Council minutes on the adoption of the decision on the signature of CETA. The undertakings refer to expediting the adoption of rules by the Joint Committee designed to reduce the costs of dispute settlement for SMEs and, 'irrespective of the outcome of the discussions within the Joint Committee', the introduction of a system of financial and technical assistance for SMEs (the latter presumably intended for EU SMEs only). According to the Court such a political commitment, linked to CETA ratification, was 'sufficient justification' on which to base a finding of compatibility with Article 47 of the Charter. 73 This reliance on a commitment to adopt further measures in order to ensure a compatible application of the agreement is something new in the Court's Opinions on international agreements. It directs the prior review procedure to an assessment of whether the CETA 'as envisaged' enables a compatible application: 'That commitment [to ensuing accessibility to SMEs] is sufficient justification, in the context of the present Opinion proceedings, for the conclusion that the CETA, as an "agreement envisaged" within the meaning of Article 218(11) TFEU, is compatible with the requirement that those tribunals should be accessible.' 74 The Court's expression gives the idea of an 'envisaged' agreement a new significance. In the past, discussion of the term has been centred on the timing of the request for an Opinion and the extent to which the details of the agreement should be available to the Court, 75 or by whom the agreement should be 'envisaged'. 76 Here, the Court is concerned with how the parties 'envisage' putting the agreement into practice. In contrast, it will be remembered, the Court in Opinion 2/13 declared incompatible with the EU Treaties a provision of the draft ECHR accession agreement that allowed, but did not require, compliance with Article 344 TFEU and the principle of autonomy; in its view, the agreement itself should have required compliance. 77 Interestingly, while mutual trust between Member States was at the basis of the reasoning in Opinion 2/13, the Court in Opinion 1/17, as already mentioned, expressly denied that mutual trust provides a basis for the CETA, using this difference to distinguish the two cases. 78 And yet the Court demonstrates a degree of trust in the implementation of the CETA, in the establishment of the ISDS tribunal and the way it will function, and in the ability of the EU institutions to insist within the CETA decision-making procedures on solutions which will be compliant with the principle of autonomy and with the Charter.

Conclusion
This short paper has examined Opinion 1/17 in its character as an Opinion delivered under Article 218(11) TFEU. The CETA Opinion follows the standard structure and adopts the analytical approach that has become characteristic of more recent Opinions, with a general statement of principle preceding the analysis of the envisaged agreement. However, the Opinion has some distinctive features. It was, somewhat unusually, requested by a Member State rather than one of the institutions, and the request was closely linked politically to the ratification of the agreement not by the EU per se, but by that Member State. The request, as the procedure requires, concerned the compatibility of the agreement with EU law, and the Opinion represents only the second time under this procedure that the Court has assessed the substantive compatibility of an agreement with the Charter of Fundamental Rights, alongside issues of institutional autonomy.
The Court was clearly concerned, in this Opinion, to demonstrate that the Union is not prevented by the principle of autonomy from developing reciprocal international relations and playing an active part in international dispute settlement. 79 Instead, the Court's focus was on the distinction between the governance of inter se relations through EU law in a context of integration and mutual trust, and the operation of external systems through international law in a context of reciprocity; on the distinction between its international action and its internal law-making, while ensuring that the former does not upset the constitutional framework which underpins the latter. In this way the Opinion builds upon Opinion 2/13 in establishing what the Court regards as the 'specific characteristics' of the EU's legal ecosystem (to use Michel Barnier's phrase 80 ), including the internal decision-making processes by which the public interest is defined and regulatory choices expressed. The principle of autonomy is certainly not limited to dispute settlement and the exclusive jurisdiction of the Court of Justice.
The separation between the CETA regime and EU law did not, however, prevent the Court from expressing a view as to the implementation of the agreement and its interpretation by a future CETA tribunal so as to protect the EU's regulatory choice. The expansion of the notion of an 'envisaged agreement' to include the way in which the parties envisage its implementation, as expressed through political statements, is an innovative aspect of the Opinion. The prior nature of the Article 218(11) TFEU review renders some degree of looking forward inevitable, especially since, often, not every detail of the envisaged agreement will be known. But the reliance on assurances external to the agreement, coupled with specific directions from the Court on the limits of the jurisdiction of an external tribunal, have the effect of reaching forward, beyond the Opinion procedure, to influence its future implementation.

Declarations and conflict of interests
The author declares no conflicts of interest with this work.