1. Introduction
In Opinion 1/17,[1] the Court of Justice of European Union (CJEU or Court) found the design of the investor-State dispute settlement (ISDS) in the Comprehensive Economic Agreement between Canada and the EU (CETA)[2] was compatible with EU law. The Court found that the agreement did not undermine the autonomy of the EU legal order. One of the criticisms of the Opinion, and of the CJEU’s jurisprudence in this area, is that the Court oscillates between different conceptions of autonomy. Academic commentary has also highlighted the ambiguous and ill-defined nature of the principle: ‘its scope has been somewhat nebulous, its limits ill-defined and its function intrinsically linked to furthering the powers of the Court of Justice’.[3] We argue that Opinion 1/17 demonstrates that the understanding of the principle of autonomy in a multi-layered order such as the EU is to a certain extent fluid and context-specific. The concept has a minimum core, consisting of the requirement that ‘the essential character of the powers of the Community and its institutions as conceived in the Treaty remain unaltered…’.[4] Beyond this, the principle remains open to interpretation and application, particularly in cases in which the Court is called upon to judge the compatibility of particular international treaties with EU law.[5] The inherent ambiguity of the concept is justified once gleaned through the lens of its function; autonomy is by its very nature a relational concept. It affirms the separate existence of a system, rule or entity vis-à-vis another system, rule or entity.[6] From this vantage point, previous expansive readings of autonomy (discussed below) can be explained on the basis of the context in which they occurred. A reading of the relevant case-law, including Opinion 1/17, through this contextual lens helps us make sense of the Court’s ‘oscillation’ between broad and narrow constructions of the principle in its practice.
Zooming in on investment arbitration, in Opinion 1/17, the nature of the agreement, between the EU and a third country, allowed for a narrow construction of the principle of autonomy and a symbiotic relationship of international and EU law. By way of contrast, in Achmea,[7] the agreement applied in the relations between EU Member States (intra-EU BIT). The different nature of the relevant treaty and thus, the highly constitutionalised setting within which the relations between the Member States took place, led the Court to apply a broad construction of autonomy.[8] Although the context-specific framing of autonomy put forward here is a helpful tool in making sense of the reasons underpinning the broad or narrow construction of autonomy in the Court’s case-law, context does not in and of itself dictate the Court’s response to each of these cases. [9]
As the Court’s development of the principle shows, autonomy is a gate to the outside world and to outside interaction with legal systems and rules. The Court itself is the gatekeeper. The principle of autonomy can be left open to allow interaction with the international legal order or it can be closed. That decision lies with the gatekeeper. As the Opinion of Advocate General Wathelet in Achmea[10] shows, autonomy does not necessarily equate to isolationism (irrespective of the setting) and the EU legal order arguably has the necessary tools to absorb any frictions that may be caused from the interaction between international and EU law. In this sense, the way in which the Court has chosen to construe autonomy each time does not necessarily say much about the principle itself, but rather it says more about how the Court chooses to construe the EU’s relationship with the wider world.
2. Evolution of autonomy
The CJEU’s jurisprudence on the autonomy of the EU legal order, particularly Opinion 1/17 and Achmea, demonstrate how the principle of autonomy is of importance, not just for the EU legal order, but also for international law more broadly. More particularly, the way that the Court frames autonomy is also relevant to the EU’s international partners – if other states wish to enter into agreements with the EU with dispute settlement mechanisms, or if the EU seeks to join international organizations with dispute settlement bodies, these agreements must ensure that they do not risk undermining the autonomy of the EU legal order. This is not a condition arising from international law, but from EU law. Yet if autonomy is a malleable concept, and applied differently depending on context, then the CJEU’s jurisprudence on autonomy does not offer suitable guidance to the drafters of those agreements to ensure they satisfy autonomy requirements. Moreover, the reaction of many commentators to the Court’s autonomy jurisprudence, especially following Opinion 2/13,[11] demonstrates how the Court’s application of the autonomy principle can give rise to surprising results. A common criticism is that the principle of autonomy, as developed through the Court’s case-law, has become an overly ambiguous and malleable concept, which makes it difficult to predict whether a given agreement conforms with EU law. Such uncertainty and unpredictability can also be seen as the EU enters into agreements in the area of investment that contain dispute settlement chapters.
The Court’s approach to the principle of autonomy may be understood as fluctuating between narrow and broad conceptions autonomy. A ‘broad’ conception of autonomy is one that requires any envisaged agreement to ensure that there is no possibility, even hypothetically, of a body other than the CJEU, interpreting EU law. It also means that any arrangements that allow for dispute settlement must ensure the ‘specific characteristics’ of the EU legal order are preserved, but it also understands that these specific characteristics include a broad and expanding range of concepts. A ‘narrow’ conception of autonomy, on the other hand, allows the EU to participate in dispute settlement bodies as long as sufficient ‘safeguards’ are put in place to protect the EU and its autonomy.[12] This narrow conception of autonomy is mainly concerned with preserving the judicial monopoly of the CJEU. As discussed below, the Court’s approach – between a narrow or broad conception of autonomy – depends on how the Court conceives of the relationship between EU law and international law.
There has been quite some debate about the conceptual origin and rationale for EU autonomy. In Opinion 1/17 the CJEU presented its understanding of the concept:
The Court does not introduce anything new here. It asserts that autonomy stems from the ‘essential characteristics’ of the EU and its law. A core part of this is that the EU and its law are not derived from any other source, from either international law or the law of the EU Member States – they have an independent existence. Understood this way, autonomy is not a relative concept at all, but absolute. An agreement cannot partly undermine the independent existence of the EU legal order. The Court’s understanding of autonomy as an absolute, binary legal concept, stands in contrast to academic commentary that often presents autonomy as a relational concept.[14] Yet these are not as contradictory as may first appear. This is because, although the Court presents autonomy as in absolute terms, the cases in which the Court deals with these issues involve the EU’s relationships and interactions with other legal orders – those of the EU Member States, or international law. As AG Bot mentions in his Opinion, ‘the preservation of the autonomy of the EU legal order is not a synonym for autarchy’,[15] observing that the principle should not be understood as requiring EU isolationism. Yet understanding what the ‘essential characteristics’ of the EU entails, and in what ways they may be undermined, does depend on the circumstances of each case, and the type of relationship that is envisaged. There may be some types of relationship, such as those with the legal order of the European Court of Human Rights, which the Court views as having greater potential for jeopardising the ‘essential characteristics’ of the EU legal order, and thus require greater scrutiny. Yet there may be instances where the relationship, such as that with the CETA Tribunal in Opinion 1/17, involves fewer possibilities for those ‘essential characteristics’ to be threatened. Thus, while the Court understands autonomy in absolute terms, it has to consider how this principle is to be applied according to the type of relationship, and the broader overall context of that relationship. This is explained in more detail below in relation to Opinion 1/17 and Achmea.
When understood as a relational concept, the Court’s oscillation between narrow and broad conceptions of autonomy, may appear more comprehensible. The obvious problem, however, is that such case-law may not give a useful guide to drafters when seeking to determine when an envisaged agreement will comply with EU law requirements. There is a tendency to understand autonomy as requiring an envisaged agreement to have certain ‘safeguards’. This means that certain institutional provisions are included, for example, to ensure that the CJEU is given the authoritative word on the interpretation of EU law. However, if the principle of autonomy is understood in a more relational fashion, the safeguards that are needed will depend on the type of relationship that is envisaged. The principle of autonomy should be understood as an overarching constitutional principle, one that is to be safeguarded in different ways, depending on the circumstances. Much of the criticism of the CJEU has been that it has been inconsistent in the application of the principle. Yet such inconsistency is understandable, and perhaps even justified, if the issues that threaten the ‘essential characteristics’ of the EU legal order also differ from case to case. In this light, while it is acknowledged that issues of predictability and legal certainty for international actors who wish to contract with the EU may still remain, the analysis provided here (hopefully) goes some way towards explaining the CJEU’s seemingly inconsistent rulings on the matter.
3. Opinion 1/17 – Reflections on autonomy and the future of ISDS
3.1. Political context of the judgment
As discussed above, the precise contours of the principle of autonomy largely remain nebulous and, thus, to a certain extent, context-specific. Statements to the effect that autonomy implies that ‘the essential character of the powers of the Community and its institutions as conceived in the Treaty remain unaltered’[16] support this proposition. Apart from a minimum core, which vaguely alludes to the preservation of the essential characteristics of the Union, the principle remains open-ended and is left to be clarified on a case-by-case basis. Against this background, this contribution zooms in on Opinion 1/17 and its wider implications.
In order to understand what was at stake, it is important to note the politically charged context in which the request for an Opinion was made. The request reached the Court amidst considerable public disquiet about the role of ISDS mechanisms in international trade agreements. More particularly, concerns have been raised (which the EU has also shared) that current ISDS mechanisms are inappropriate for disputes involving States due to lack of guarantees that arbitrators are independent, lack of consistency and foreseeability of awards, the high cost of the proceedings, as well as lack of transparency.[17] In this light, the EU has made concrete efforts to reform the existing ISDS model. First, in the context of CETA, a new model for resolving disputes between foreign investors and States (and the EU) was introduced, namely the Investment Court System (ICS) model.[18] Under this model, disputes arising under CETA will be heard by a permanent tribunal composed of independent and publicly appointed members (CETA tribunal) whose decisions are subject to an appellate body. It should be noted that the EU has also been promoting the establishment of a single multilateral investment court building on the ICS model in the context of the UNCITRAL. The CETA already contains a reference to the effect that: ‘[u]pon the establishment of such a multilateral mechanism, the CETA Joint Committee shall adopt a decision providing that investment disputes under this section will be decided pursuant to the multilateral mechanism and make appropriate transitional arrangements.’[19]
However, not everyone was convinced. A number of civil society actors and academics have objected to the ICS mechanism included in CETA as they perceive it to be in tension with the autonomy of the EU legal order.[20] Concerns regarding the merits of CETA were also raised by Belgian regional parliaments – which led Belgium to ask for an Opinion by the Court regarding the compatibility of CETA with EU law pursuant to Art. 218(11) TFEU. In the meantime, the CJEU’s ruling in Achmea to the effect that ISDS provisions in intra EU-bilateral investment treaties (intra EU-BITs) are incompatible with the principle of autonomy of EU law[21] led to widespread speculation about whether the Court would follow a similar line of argumentation in relation to ISDS mechanisms provided for under investment treaties with non-EU countries. Would the Court also find that the ICS mechanism under CETA is incompatible with the principle of autonomy of EU law given its long-standing concerns about protecting the autonomy of the EU legal order in its previous case-law?
The political stakes were very high; a finding of CETA’s incompatibility with the principle of autonomy would have effectively signalled the death knell for the EU’s efforts of modernizing and promoting investment protection. As it is, the judgment removes significant hurdles for the ratification of CETA by Member States as well as the ratification of a number of other agreements containing similar ISDS provisions (such as the agreements with Singapore and Vietnam) and it significantly bolsters the EU negotiating position in the context of the ongoing UNCITRAL negotiations on ISDS reforms. Of course, it is not unusual for an Opinion procedure to involve a political dimension. Opinion 2/13, for example, took place in a context where the Court’s decision would have significant repercussions. At the same time, it needs to be highlighted that the legal and factual circumstances surrounding the present Opinion were substantially different. As it will be shown below, the ‘autonomy safeguards’ included in CETA imply that the increasingly high threshold of autonomy put forward by the Court in its case-law has not fallen on deaf ears. On the contrary, the inclusion of these safeguards attests to the political significance of the case. The extremely careful couching of the terms of the agreement means that its drafters took seriously into account the previous case-law of the Court on autonomy – and that they took great pains to minimize the potential for jeopardising the ‘essential characteristics’ of the EU legal order.
3.2. The judgment
The Court began its analysis of the compatibility of CETA’s ISDS provisions with EU law by recalling its settled case-law that an international agreement providing for the establishment of an international court whose decisions are binding on the EU is, in principle, compatible with EU law provided that the autonomy of the EU legal order is respected.[22] The autonomy of the EU legal order is guaranteed by the judicial system established under the Treaties, which provides for the exclusive jurisdiction of the CJEU to give the definitive interpretation of EU law – a feature which is further supported by the preliminary reference procedure envisaged under Art. 267 TFEU.[23] The Court noted from the outset that the fact that CETA’s ICS stands outside the EU judicial system does not mean, in and of itself, that it adversely affects the principle of autonomy.[24] The reciprocal nature of international treaties and the need to maintain the powers of the EU in international relations entail that the EU may conclude an agreement creating an international court that is not subject to the interpretations of that agreement given by the domestic courts of any of the parties.[25] The Court continued by stating that while EU law does not preclude, in principle, the creation of the CETA tribunal, the principle of autonomy dictates the examination of two points: a) that the CETA tribunal does not have the power to interpret or apply EU rules other than the CETA provisions having regard to the rules and principles of international law applicable between the parties; and b) that the CETA tribunal does not have the power to issue awards having the effect of preventing the EU institutions from operating in accordance with the EU constitutional framework. In other words, the Court examines whether these provisions ensure that the CETA tribunal does not have the power to influence EU law indirectly.[26]
Against this backdrop, the Court proceeded to examine the text of the CETA. As regards the first point, the Court highlighted that the jurisdiction of the CETA tribunal was confined to the interpretation and application of the relevant CETA provisions in accordance with the principles and rules of international law applicable between the parties.[27] It was also important that Art. 8.31.2 CETA expressly states that its bodies will not have jurisdiction ‘to determine the legality of a measure, alleged to constitute a breach of this Agreement, under the domestic law of a Party.’ These provisions guarantee that the Tribunal would not have the power to interpret or apply EU law.[28] This guarantee is, according to the Court, what distinguishes CETA from other agreements that it had found problematic in the past (such as the Netherlands-Slovakia BIT in Achmea).[29] The Court further distinguished between CETA and intra-EU BITs by underscoring that the principle of mutual trust, which played a major role in its line of reasoning in Achmea, does not apply to agreements between the EU and third states.[30] In this context, the Court also examined certain other ‘autonomy safeguards’ included in Art. 8.31 CETA and found them sufficient to guarantee the principle of autonomy of the EU legal order. These include provisions to the effect that in determining whether a measure violates the agreement, the tribunal may only consider domestic law as a ‘matter of fact’ following the prevailing interpretation given by domestic courts and that domestic courts are not bound by the meaning given to their domestic law by the tribunal.[31]
As regards the second aspect, the Court found that the awards issued by the CETA tribunal would not have the effect of preventing the EU institutions from operating in accordance with the EU constitutional framework. In particular, the Court addressed here the question of whether the creation of the CETA tribunal could have an adverse impact on the Union’s right to define the level of protection of public interests under EU law – especially in the context of examining a defence put forward by the Union in response to a breach of the substantive protections under CETA.[32] The Court conceded that the principle of autonomy precluded the conclusion of an agreement:
However, the Court was satisfied that CETA provides sufficient safeguards in this respect. It contains various clauses expressly confirming that the CETA tribunal does not have jurisdiction to declare incompatible with CETA the level of protection of a public interest established under EU law as well as guaranteeing the right to regulate in the public interest.[34]
3.3. Implications of the judgment
The Opinion demonstrates how the application of the principle of autonomy is to a large extent fluid and context-specific. By way of contrast to previous autonomy case-law, both the Advocate General[35] and the Court put great emphasis on ‘the reciprocal nature of international agreements’ as well as on ‘the need to maintain the powers of the Union in international relations’.[36] This express reference to reciprocity illustrates the Court’s acknowledgement of the intense political overtones of the case; a finding to the effect that CETA were incompatible with the principle of autonomy would have been a rather bold move on the part of the Court undermining the efforts of the EU to modernize investment arbitration in the context of UNCITRAL.[37] This is not the first time that the Court relied on an argument of ‘reciprocity’. In the context of denying the direct effect of the WTO Agreements, the Court also took into account that the agreements were ‘based on the principle of negotiations undertaken on the basis of ‘reciprocal and mutually advantageous arrangements’.[38] In that context, direct effect could cause a lack of reciprocity vis-a-vis other WTO members and thus, it could entail the non-uniform application of WTO law. However, it needs to be stressed that the situation at hand is different. That line of case-law concerned the effects of an international agreement within the EU legal order, whereas the present case concerned the compatibility of an external dispute settlement mechanism with EU law.
The Court carefully distinguished between the specific legal and policy context of the dispute and the one which led to the Achmea ruling. In Achmea, the Court found that the intra-EU context of the agreement was problematic to the extent that questions about EU law could arise before arbitral tribunals which themselves do not form part of the EU judicial system. Since they cannot ask the CJEU preliminary questions, this would have an adverse effect on the principle of mutual trust as well as on the preservation of the specific characteristics of the EU legal system and its autonomy.[39] Advocate General Bot succinctly summarised the line of reasoning in Achmea by stating that the Court’s approach was:
The fact that CETA is an agreement between the EU (and its Member States) and Canada means that the concerns raised in relation to intra-EU BITs are not applicable here. As the Court noted in its judgment, the principle of mutual trust is not applicable in the relationship between the EU and Canada, and thus, CETA cannot adversely affect the principle in question.[41] Furthermore, the fact that the CETA tribunal, in contrast to the one in Achmea, can only take into account EU law as a matter of fact means that the exclusive prerogative of the CJEU to provide binding interpretations of EU law is not threatened, thereby minimizing concerns regarding the potential impact of the agreement on the principle of autonomy.[42]
Opinion 1/17 also confirms that although the context of a dispute is a helpful tool in making sense of the reasons underpinning the construction of autonomy in the Court’s case-law, context does not in and of itself dictate the Court’s response. As discussed above, autonomy is a gate to the outside world and to outside interaction with legal systems and rules. By way of contrast to a line of case-law where autonomy was couched in abstract and uncompromising terms – thereby employing autonomy as an instrument to highlight the separateness and primacy of EU law over international law, the point of departure in Opinion 1/17 is different. By emphasising the reciprocal nature of international agreements, the Court showed its willingness to envision a symbiotic relationship between CETA and EU law. The safeguards put in place can absorb any tension that may arise between the two legal regimes.
The Court’s vision of the EU legal order’s relationship with international law (antagonistic vs. symbiotic) lies at the heart of its (narrow or broad) construction of the principle of autonomy in different cases. This is further bolstered when one takes a closer look at the safeguards in place to secure the autonomy of EU law in Opinion 1/17. Here, the Court made sure to distinguish CETA from previous case-law: the CETA tribunal would not have jurisdiction to interpret and apply EU law (as opposed to Opinion 1/09); the tribunal would not be created on the basis of an agreement between Member States and thus, the principle of mutual trust would not be relevant (as opposed to Achmea); and the tribunal’s (incidental) interpretations of EU law would not be binding upon the EU and on Member States (as opposed to Opinion 1/91).
There were issues that (arguably) required further scrutiny. In Opinion 2/13 and in Achmea, even a hypothetical interference with the EU legal order was enough for the Court to find an incompatibility. In the context of CETA, this abstract possibility is not entirely eliminated. Even if the CETA Tribunal is bound to follow the CJEU’s interpretation of EU law, such an interpretation may simply not yet exist since investors can bring claims on the basis of recently enacted legislation which may have not yet been interpreted by domestic courts.[43] Thus, there is still an abstract possibility that the CETA tribunal may provide its own interpretation of EU law. Such an interpretation would not be binding at the EU law level, however, since the tribunal may consider domestic law only ‘as a matter of fact’ and the domestic courts are not bound by the meaning given to their domestic law by the tribunal. This differs from the situation in Opinion 2/13, whereby a decision of the ECHR would have been binding on the EU and the Member States. Although autonomy may mean different things in different contexts, this shows how much depends on the Court’s own construction of the relationship between EU and international law in a given case. In Opinion 1/17, the extra EU-BIT nature of the agreement helped steer the Court towards envisaging a symbiotic relationship between international and EU law. Had the point of departure been different, the Court would likely have probed further, moving towards a broad conception of autonomy.
This point becomes even clearer if one takes into account the different reasoning followed by the Court and AG Wathelet respectively in Achmea. In Achmea, the intra EU-BIT nature of the agreement arguably played a significant role in the Court’s ruling. At the same time, the Court’s finding was not a foregone conclusion. The Court in that case chose a broad and formalistic line of argumentation based on the autonomy of EU law – without embarking on a detailed examination of Art. 344 TFEU to assess whether that Article allowed for investor-State arbitration.[44] In this sense, autonomy was seen in Achmea as a fence emphasising the separateness of the EU and the international legal order. However, the opinion of AG Wathelet in Achmea shows that there is an alternative way of reading autonomy – even in that particular context – namely as a bridge connecting the two legal orders. AG Wathelet argued in Achmea that arbitral tribunals established by intra EU-BITS meet all the necessary requirements in order to be considered as ‘courts’ or ‘tribunals’ within the meaning of Art. 267 TFEU and that since they are common to the two Member States they are bound by Art. 267 as far as the application and interpretation of EU law is concerned. Furthermore, according to the AG, the existence of these arbitral tribunals does not violate Art. 344 TFEU since they are required to respect principles of EU law, and failure to do so would result in the liability of Member States for damages and in enforcement actions by the Commission/other Member States. In this sense, the AG’s point of departure was that of a symbiotic relationship between EU law and investment law, and this allowed him to construct autonomy differently than the Court. Instead of formalism and distrust, the AG put forward the idea that international investment law and EU law can harmoniously co-exist, and that EU law contains enough safeguards (preliminary reference procedure, the principle of State liability) that its autonomous nature will not be eroded through interaction with another legal order.
The ruling is also of particular significance for the future development of the principle of autonomy. By construing autonomy narrowly and by avoiding the ‘maximalist overtones’[45] permeating Opinion 2/13 and Achmea, the Court brought the concept closer to its conceptual origins (and thus more in line with its earlier case-law thereon). Opinion 1/17 reflects the idea that autonomy is a mechanism for ensuring that the EU and its institutions are not bound by a particular interpretation of EU law stemming from a body that stands outside the EU judicial system. It is not merely a mechanism for ensuring the judicial monopoly of the CJEU.
More broadly, the ruling has important consequences for the EU’s participation in international dispute settlement procedures – and, in general, for the EU’s role as a global actor. Both Opinion 2/13 and Achmea were strong assertions of autonomy that cast doubt on the EU’s ability to participate in the wider international legal system. The narrow construction of autonomy in Opinion 1/17 coupled with the fact that the Court accepted both in theory and in practice that EU law and international law can co-exist harmoniously shows a considerable degree of openness towards international law. Ultimately, the judgment portrays the EU as a more self-confident and outward-looking global actor.
4. Conclusion
Although the Court is described as moving between narrow and broad conceptions of autonomy, it has never really departed from the original conceptual basis, that is, the separate and independent existence of the EU legal order. The oscillation between narrow and broader conceptualisations of autonomy can be explained by the context and how the Court views the relationship between EU law and international law in each case. While Opinion 1/17 brought the principle of autonomy closer to its conceptual origins and to earlier case-law, the Court’s approach still stands in contrast to the broad conceptions adopted in Achmea and Opinion 2/13. However, this also comes at the expense of legal certainty and clarity. Moreover, the EU will not always be in a position to include safeguard provisions as carefully drafted as those included in the CETA agreement. The Court’s application of the principle has broader significance for the EU’s relationship with international law. As the EU continues to become integrated in the international legal order, especially through its participation in international dispute settlement mechanisms, the Court’s jurisprudence on autonomy will continue to shape the nature of this relationship. Autonomy will continue to evolve and find new forms of application depending on the circumstances and political context.
* Senior Researcher in EU and International Law T.M.C. Asser Institute, Academic Co-ordinator of CLEER.
** Lecturer in Law, City Law School, City, University of London.
[1] Opinion 1/17 EU:C:2019:341.
[2] ‘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 11/23.
[3] P Koutrakos, ‘The Autonomy of EU Law and International Investment Arbitration’ (2019) 88 Nordic J Intl L 1, 49.
[4] Opinion 1/00 EU:C:2002:231 para 12.
[5] The principle of autonomy also arises in other areas of EU law, such as deciding the effect of international law within the EU legal order. See J Odermatt, ‘The Principle of Autonomy: An Adolescent Disease of EU External Relations Law?’ in M Cremona (ed), Structural Principles in EU External Relations Law (Hart 2018) 305.
[6] ‘Autonomy would seem to be, somewhat counter-intuitively perhaps, a nonautonomous, relational concept: one cannot be autonomous in isolation, or rather when in isolation the concept of autonomy does little work.’ J Klabbers, P Koutrakos, ‘Introduction: An Anatomy of Autonomy’ (2019) 88 Nordic J Intl L 1, 4.
[7] Judgment in Slovak Republic v Achmea BV, Case C‑284/16 EU:C:2018:158.
[8] ‘Autonomy is understood in broad and uncompromising terms, and compliance with it may not envisage any alternatives for safeguarding the essential features of the EU’s legal order’ Koutrakos (n 3) 56.
[9] See V Moreno-Lax, ‘The Axiological Emancipation of a (Non-) Principle: Autonomy, International Law and the EU Legal Order’ in I Govaere, S Garben (eds), The Interface Between EU and International Law: Contemporary Reflections (Hart 2019) 71, where she notes that: ‘There is no a priori determinism of what autonomy entails’.
[10] Opinion of Advocate General Wathelet, Slovak Republic v Achmea BV, Case C‑284/16 EU:C:2017:699.
[11] Opinion 2/13 EU:C:2014:2454.
[12] T Locke, ‘The Future of the European Union’s Accession to the European Convention on Human Rights after Opinion 2/13: Is It Still Possible and Is It Still Desirable?’ 11 (2015) Eur Constitutional L Rev 239, 243: ‘A narrow conception of autonomy, such as this, is appropriate as it serves the legitimate purpose of protecting the integrity of the EU law while retaining the EU’s capacity as an external actor’.
[13] Opinion 1/17 (n 1) para 109. Emphasis added.
[14] See B de Witte, ‘European Union Law: How Autonomous is its Legal Order?’ (2010) 65 Zeitschrift für öffentliches Recht 141, 142. J Odermatt, ‘When a Fence Becomes a Cage: The Principle of Autonomy in EU External Relations Law’ EUI Working Papers 2016/07, 1 ‘Autonomy, however, is not absolute. It does not so much describe an absolute quality of an entity, but the relationship of that entity with others, and in particular the ability of that entity to define this relationship’.
[15] Opinion 1/17, Opinion of AG Bot, ECLI:EU:C:2019:72 para 48.
[16] Opinion 1/00 (n 4) para 12.
[17] SD Franck, ‘The Legitimacy Crisis in Investment Treaty Arbitration: Privatizing Public International Law through Inconsistent Decisions’ (2005) 73 Fordham L Rev 1521.
[18] CETA, Chapter 8 section F.
[19] Art 8.29 CETA.
[20] L Ankersmit, ‘Investment Court System in CETA to be Judged by the ECJ’ Eur Law Blog (31 October 2016) <http://europeanlawblog.eu/2016/10/31/ investment-court-system-in-ceta-to-be-judged-by-the-ecj/>.
[21] Case C‑284/16 (n 7) paras 59-60.
[22] Opinion 1/17 (n 1) paras 106-107.
[23] ibid para 111.
[24] ibid para 115.
[25] ibid para 117.
[26] ibid paras 118-119.
[27] ibid para 122.
[28] ibid paras 122, 133.
[29] ibid para 126.
[30] ibid paras 127-128.
[31] ibid paras 130-131.
[32] ibid paras 137, 149, 150.
[33] ibid para 150.
[34] ibid paras 130-131.
[35] Opinion of Advocate General Bot (n 15) paras 72-90.
[36] Opinion 1/17 (n 1) para 117.
[37] P Koutrakos, ‘More on Autonomy – Opinion 1/17 (CETA)’ (2019) 44 Eur L Rev 293, 293.
[38] Judgment in Germany v Council, C-280/93 EU:C:1994:367 para 106.
[39] Case C-284/16 (n 7) paras 30-32, 34, 58.
[40] Opinion of Advocate General Bot (n 15) para 105.
[41] Opinion 1/17 (n 1) paras 128-129.
[42] ibid para 131.
[43] M Gatti, ‘Opinion 1/17 in Light of Achmea: Chronicle of an Opinion Foretold?’ (2019) 4 Eur Papers 1, 9-10.
[44] S Gaspar-Szilagyi, ‘It Is Not Just About Investor-State Arbitration: A Look at Case C-284/16 ACHMEA BV’ (2018) 3 Eur Papers 357, 365.
[45] Koutrakos (n 3) 60.
