Outer Scrutiny in International Law
CHAPTER I – OVERVIEW OF THE Ph.D.
introduction
This introductory chapter aims to delineate the background, scope, aims, and methodology used in this Ph.D. Section 2 gives a brief overview of the topic of this thesis – the definition of ‘outer scrutiny’ of international courts and tribunals, and how its interpretation by the Court through the lenses of the ‘autonomy of EU law’ may impact on the identity of the EU as a promoter of the rule of law and judicial protection in international law. Section 3 sets out the aim and relevance of this work. Considering the large body of literature that already exists on the topic of autonomy of EU law, the section also attempts to highlight how this Ph.D. intends to add to the current doctrinal debate. A brief account of the scope, methodology, and terminology used in this work is given in section 4. The conclusions in section 5 provide an outline of the remainder of the Ph.D.
the issue: outer scrutiny, judicial protection, and the undefined notion of autonomy in the EU
This Ph.D. builds on the idea that in today’s world, characterized by an increasing number of transnational relations and a central role for non-state entities as rights carriers, the rule of law no longer is a concept linked merely to national law. In this context, ‘outer scrutiny’ of international courts and tribunals over states, international organisations (including the EU), and non-state entities alike is an essential component of an idea of the rule of law that crosses borders: while these dispute settlement mechanisms are by no means the only instrument to promote adherence to international norms and resolve international disputes, they undoubtedly possess an increasingly important central role to ensure compliance with the law over the use of force and arbitrariness in the international relations. The notion of outer scrutiny of international courts and tribunals as a tool to promote the rule of law across borders is arguably part of the DNA itself of the European Union: the Union is the product of a global movement commenced at the end of the 19th century which saw in the creation of supranational institutions the solution to enforce the rule of law over the rule of power. Is with that aim in mind that the 1951 treaty of Paris, establishing the European Coal and Steel Community (ECSC) placed the Franco-German production of coal and steel under the control of a supranational organization provided with its own institutions, including a Court of Justice. The role of the ‘outer scrutiny’ of the system of European courts became increasingly more important in the Community set up a few years later by the treaty of Rome. Notably, not only the (at the time) ECJ was one of the earliest examples of international judiciary set up to ensure compliance with ‘European’ rule of law standards. The ECJ also undoubtedly was the most effective dispute resolution mechanisms of its times. A variety of factors – including its compulsory jurisdiction, institutional ties with national courts of the Member States through the preliminary reference procedure, and the favourable historical and political environment – allowed the CJEU an unseen degree of independence by the Member States, and made it stand out from most of its international peers, like the International Court of Justice (ICJ). Among these, one of the most significant was that it conferred standing to individuals before a network of European courts. By breaking their monopoly to demand compliance with the law, direct (through direct actions) and indirect (through the preliminary reference mechanism) individual standing undoubtedly contributed to the success of the ‘outer scrutiny’ by the CJEU of Member States’ compliance with EU law and became an element of the identity of the EU. Today, the CJEU is one among many other international dispute settlement mechanisms able to effectively exercise outer scrutiny across borders and safeguard the rule of law, and no longer the most effective one. A growing network of active international dispute settlement mechanisms enforces obligation, makes individuals and entities accountable, and effectively ensures that the rule of law is complied with not only across the 27 Member States, but across continents, in a broader number of fields than the CJEU, and through mechanisms unknown to the European judiciaries. The legitimacy, independence, and effectiveness of these international dispute settlement mechanisms allows them to perform functions that go beyond the settlement of disputes and are much closer to a form of global governance: these encompass the review by courts of acts of national and international authorities, rendering judicial decisions that create basic source material for international law, contributing to the effective development and enforcement international law internationally important fields of international law, altering state policy by circumventing domestic legal and political barriers and to create legal change across borders, creating communities of actors pursuing the same agenda on a global scale, and diminish political control over international law. Among these international dispute settlement mechanisms ISDS arbitration stands out. The particular independence of this mechanisms from states, the fact that it puts individuals on (almost) the same footing as states, the effective mechanisms of enforcement of international law standards, have indeed earned ISDS arbitration the definition of a particularly powerful tool of global governance, or a “species of Global Administrative Law.”
Outer scrutiny of these new powerful dispute settlement mechanisms creates however a dilemma for the EU: on the one hand, as a commitment to compliance with the rule of law and individual protection, outer scrutiny arguably has become part of the common identity of “peoples of Europe” in the EU treaties On the other hand, it entails some cession of sovereign powers on the part of the EU, and requires compromises with some of its essential features, including the operation of its system of judicial protection. While the EU treaties and the EU external policy seem to resolve this impasse with an approach of principled openness towards outer scrutiny of these dispute settlement mechanisms, the CJEU has in the years adopted a more ‘protectionist’ stance and deemed the outer scrutiny of a large number of international courts and tribunals incompatible with EU law on grounds of the judge-made concept of ‘autonomy’ of EU law. Among the reasons adduced by the Court to justify its approach, the safeguard of the ‘judicial protection’ through EU courts rather than international tribunals has notably played a pivotal role. On the one hand, the CJEU is well-aware of the importance of its direct relationship with individuals and national courts for the success of the European project and appears reluctant to share it with others. On the other hand, the Court also seems to work on the historic assumption of a ‘special’ protection that individuals possess in the EU and through its courts, as opposed to international law, and appears determined to preserve it against any dangerous external interactions. That approach to international law was well summarised by Eeckhout in relation to the CJEU case in Kadi, as “Luxembourg is not Texas – or Washington DC.” Such a gap between what the EU is in the face of the EU treaties and in the light of the policy choices of the EU Institutions, and how the CJEU in practice interprets the autonomy of EU law carries with it the risk of damaging its role in international law and delegitimizing its acts. Openness of the CJEU towards the outer scrutiny of these international courts and tribunals is today no longer only advisable from a perspective of coordination of judicial functions, but a requirement to ensure that the EU lives to its identity as a champion of the rule of law internationally and as a leader in its development.
aim and relevance: redefining autonomy to reframe the role of the eu as a global leader
This Ph.D. addresses the question of what the Court understands for ‘autonomy of the EU’ and how that understanding fits with the constitutional identity of the EU as a promoter of the rule of law internationally through outer scrutiny of international dispute settlement mechanisms. The work takes as a case study the approach of the Court to arbitration, particularly ISDS arbitration. The work firstly looks at the general understanding of the CJEU of ‘autonomy of EU law.’ It does so by developing a theoretical framework for the use of autonomy by the Court that sets out its origin, legal nature, and development. That theoretical framework also includes a working definition of autonomy of EU law that will then be used throughout the rest of this work. The Ph.D. then moves on to look at the increasingly importance of international courts and tribunals in upholding the rule of law internationally and how that role is reflected in a position of principled favour towards outer scrutiny of international courts and tribunals in the EU treaties and on the part of the Commission and the Council. It is argued that a gap exists between the relationship between autonomy and outer scrutiny in the face of the EU treaties and in the policy approach of the EU to external relations and its implementation by the Court. That gap is regarded as entailing numerous damaging consequences for the role and legitimacy of the EU, and the Court itself, in the external relations. On that basis, this work theorises a new, more constitutionally compatible, paradigm of autonomy of EU law that can be applied to outer scrutiny. The Ph.D. further makes some practical suggestions on how the CJEU could implement that paradigm in its decision-making activity. The ultimate aim of this thesis is to provide constructive academic criticism on the case law of the Court on autonomy and outer scrutiny of international courts and tribunals. It does so by not only evidencing its pitfalls, but also by setting out the coordinates for a new paradigm of autonomy that may effectively contribute to advancing the European project of the “ever closer union.” The gist of this work is that it is through enhanced openness towards outer scrutiny – rather than uncompromising rejection – that the EU will fulfil its constitutional nature and policy aims to become a protector and advancer of the rule of law and judicial protection internationally and fully unleash its potential as a leader and a mover in international law. The arguments developed in this thesis build on three large bodies of existing scholarship. The first one covers the academic works on the topic of the ‘autonomy’ of EU law. The existing literature and academic research however mostly address autonomy in a sectorial manner, i.e., in relation to specific decisions of the Court. This Ph.D. instead engages with these works by attempting to take up the challenge to fill some of the remaining gaps by, firstly, developing a general theoretical framework for autonomy as it is understood by the Court, which includes its nature, aims, development, and a definition. Secondly, this works wants to reframe autonomy within the constitutional identity of the EU as it emerges from the EU treaties. It does so by looking at autonomy of EU law in the international relations from the perspective of implementing the values and aims of the EU treaties that define the Union as a supporter of the international rule of law through outer scrutiny. This work also attempts to add to the existing literature by setting out the broad lines of a new theoretical paradigm of autonomy based on dialogue and openness with international courts and tribunals that fulfils the identity of the EU, as well as by providing some practical guidance for the CJEU to implement it in its future decisions. The second body of existing scholarship this Ph.D. relies upon is the one revolving around the general relationship between EU law and international law rules, mechanisms, and bodies that, directly or indirectly, influence the development of EU law. That notably includes the relationship between the EU with the ECtHR and other international human rights bodies; the use of the Court of general principles of international law and customary international law; and many others. As this body of literature is only indirectly relevant to the arguments made in this work, when relying upon it this Ph.D. makes references to the more developed thinking of other authors. The contribution that this work aims to make to this second body of literature is more modestly linking the arguments made therein to the broader academic discourse on autonomy and outer scrutiny. In order to analyse the evolution of the role of international courts and tribunals, and specifically ISDS, this Ph.D. also looks at the existing literature on the evolution of the role of international dispute settlement mechanisms to uphold the rule of law. The contribution that this work intends to make to this third body of literature is limited to a reflection on the development of the role of the CJEU in the EU legal system compared to international dispute settlement mechanisms and how the EU legal system may benefit from enhanced dialogue with them.
scope, methodology, and terminology
This work will focus on the relationship between autonomy and the role of outer scrutiny of international courts and tribunals as elements of the identity of the EU. It should be noted from the outset that this Ph.D. uses indistinctively the expressions ‘autonomy of EU law’ and ‘autonomy of the EU legal system.’ While the latter expression seems to this author as more accurately describing the doctrine of autonomy in the specific context of the EU, for it refers to the constitutional characteristics of the entirety of the EU legal system rather than only of its laws, it is acknowledged that the two expressions are normally used interchangeably by the wider doctrine. The arguments developed in this thesis are specifically targeted at the CJEU. However, the overall conclusion that a concept of ‘autonomy of EU law’ that fulfils the aims and values of the EU must entail dialogue and collaboration with other international courts and tribunals enforcing the rule of law internationally, makes some of these arguments also relevant for other dispute settlement mechanisms directly or indirectly exercising outer scrutiny over EU law. In addition to this, the critique against the reform process of ISDS arbitration being carried out by the Commission discussed in chapter […] is also addressed to other EU Institutions and international actors that support the transition towards a court-like system of settlement of investment disputes. The theoretical framework underpinning this Ph.D. is placed beyond the debate on the legal nature of the EU and assumes a strictly constitutionalist perspective. This thesis also builds on the constructivist strand of international relations theory, which views the law as a reflection of ideas, norms, and identities, including in the areas of foreign policy and global governance. That is particularly evident in the way this Ph.D. draws on foreign policy objectives and values to discuss the evolving confines of the identity of the EU and to establish the ‘normative basis’ for its foreign policy and ‘distinctiveness’ in international law. The methodology used for this Ph.D. is normative, descriptive, and encompasses a critical analysis. ‘Normative’ because the thesis analyses the legal framework presiding to the development and implementation of the autonomy of EU law, specifically in relation to the identity of the EU as a promoter of the rule of law in the international relations the outer scrutiny of international dispute settlement mechanisms. That framework primarily consists of the EU treaties, but also encompasses other sources of primary and secondary EU law, including the Charter of Fundamental Rights of the EU (CFREU), general principles of EU law, EU regulations and directives, Council mandates for negotiations of international agreements, and the text of existing EU trade and investment agreements. While this thesis focuses specifically on EU law, other important international legal sources referred to in this Ph.D. are: international treaties – particularly international investment agreements, – customary international law, international arbitration rules, and the drafts of future investment agreements, especially those currently being negotiated by the European Union. This Ph.D. is also ‘descriptive,’ for it describes how the law is applied in the practice of the EU Institutions, particularly by the CJEU. The analyzed decisions of the CJEU all attain to the implementation by the Court of the concept of autonomy in relation to the outer scrutiny conducted over the EU legal order by international dispute settlement mechanisms, particularly arbitral tribunals. Where relevant, those decisions are compared with the jurisprudence of national and arbitral tribunals, including ISDS tribunals, WTO panels, the European Court of Human Rights, and the International Court of Justice. It should be noted that the judgments of the CJEU and of all the other decision-making bodies included in this Ph.D. are quoted from their English versions, except for cases in which there is an interesting nuance in the original language. It should also be noted that the limited space allowed in this work constrained the number of cases of the CJEU that could be analysed in full to define its approach to autonomy and outer scrutiny. The thesis thus relies on the research technique of ‘purposive sampling’ – a technique based on the assumption that, from a small sample, one may judge the whole piece. The methodology used for this Ph.D. was then completed with a ‘critical analysis’ of the practice of the application of the law by the CJEU. For the purpose of providing clarity or exemplifying certain concepts, that critical analysis occasionally also includes reflections on the historical development of the areas of law and case law discussed. However, it should be noted that this Ph.D. does not attempt to provide a full chronological analysis of the legislation and cases discussed, which should be looked at from an a-temporal perspective for the only purpose of highlighting certain changes in the approach of the legislators or the Court. The research underpinning this thesis was mainly conducted through desktop search, and specifically through the use of legal doctrinal and case law databases, material in various libraries, and reports of international bodies, both in the fields of EU and international law.
It should also be mentioned that this research was influenced by the year I spent working at the Permanent Court of Arbitration, where I worked on a variety of intra-EU BITs cases, especially relevant for the case study of this Ph.D. Notably, my time at the Permanent Court of Arbitration coincided with the post-Achmea phase, where the investment tribunals in the cases I worked on were being invested with the decision of the ‘Achmea objections’ raised by respondent EU Member States. Those objections were aimed at asking arbitral tribunals to reconsider or decline their jurisdiction on the grounds of the determination of the CJEU in Achmea. While those cases will not be discussed here for reasons of confidentiality (most of them are not in the public domain), they profoundly affected the underlying arguments of this work, especially the perspective of arbitration as a tool to afford effective judicial protection to individuals. The time at the Permanent Court of Arbitration afforded me, a lawyer educated in a civil law jurisdiction and for whom the ways of European Union law always had been an unquestionable assumption, the opportunity to understand the different ways of international arbitration and the standpoint of arbitral tribunals on the EU. During my year in The Hague I also participated in the drafting of a new set of arbitration rules for business and human rights. That experience gave me an in-depth insight in the potential of arbitration as a tool to ensure effective judicial protection and pursue the rule of law internationally, on a scale and through instruments unknown to EU law. As it will hopefully become evident from the remainder of this work, while these experiences have reinforced my belief in the great merits of the European project, they also have broadened my understanding of the potential and limits of both systems of EU law and international arbitration, which resulted in the addition of a more ‘practical’ gloss to this research. That practical gloss emerges most clearly in the final chapter of this work, where I develop a new paradigm of autonomy, as well as in the underlying message of the Ph.D. that it is in compromise and synergy between like-minded systems built around the protection of the rule of law through access to remedies internationally that lay their advancement and success. A final caveat is necessary. The law and case law underpinning the subject matter of this Ph.D. are in constant evolution. Almost every month, and sometimes weekly, new developments arise that may influence the cases and arguments discussed in this work. Therefore, this Ph.D. should be considered up to date at the date of submission.
conclusions: the outline
This Ph.D. has a ‘pyramidal’ structure, as it moves from defining the abstract understanding of ‘autonomy’ of the Court, to identifying the systemic constitutional issues with its application to the relationship with international courts and tribunals carrying out outer scrutiny on the EU, to discussing the consequences of that application in terms of rule of law and judicial protection, and, finally, to developing a new paradigm of autonomy more compatible with the identity of the EU set out in the EU treaties, with some practical tips to implement it. This thesis is organised as follows. Part I of this Ph.D. sets out a ‘theory of autonomy of EU law.’ The aim of Part I is to understand the origin, legal nature, and development of the concept of autonomy in the EU legal order as it is understood by the Court. For that purpose, Part I firstly sets the scene for this Ph.D. by looking at the use and content of autonomy in legal philosophy and in relation to collective entities. Part I then moves to specifically explore the constitutional origin and doctrinal nature of autonomy of EU law as understood by the CJEU, focussing specifically on its interface with the principle of effective judicial protection and on the role inter-judicial dialogue with national and international courts in its development and operation. Finally, Part I develops a working definition of autonomy of EU law that will be relied upon for the remainder of this Ph.D. Those considerations form the basis for a comparison with the autonomy of EU law by the Court in the following part of the Ph.D. Part II of this Ph.D. focuses on ‘autonomy and outer scrutiny.’ After distinguishing between ‘internal’ and ‘external’ autonomy of EU law, Part II clarifies that this work focuses on this second aspects, i.e., the constitutional relationship of the EU with international law. The thesis then moves to analyse the important role that outer scrutiny of international courts and tribunals possesses today in upholding the rule of law internationally and how that important role is today reflected in the EU treaties and in the external policy of the EU. Specifically, the Ph.D. maintains that while entailing certain limitation of the sovereign powers of the EU, outer scrutiny of international courts and tribunals is today part of the commitment to multilateralism of the EU and even an element of its own identity as a promoter of the rule of law in the international relations. The principled favour of openness towards these dispute settlement mechanisms both in the EU treaties and on the part of the Commission and Council, it is argued, should thus be also reflected in the case law of the CJEU (Chapter III). Part II of the Ph.D. then moves on to contrast the described approach of the EU treaties with the understanding of autonomy from outer scrutiny adopted by the CJEU, taking as a case study arbitral tribunals and particularly ISDS arbitration (Chapter IV). Part III of this Ph.D. concludes that the approach to outer scrutiny displayed by the CJEU sits uncomfortably with the identity of the EU and its policy approach to the international relations, and thus needs to be reformed. The gap between what autonomy from outer scrutiny is in the face of the treaties and the EU external policy and its interpretation by the Court risks undermining this latter’s legitimacy and effectiveness in the external relations, with overall negative consequences for the European project. That is especially true for the instrument of ISDS arbitration (Chapter V). The thesis thus suggests a change of paradigm of autonomy from outer scrutiny on the part of the Court, defined in this work as ‘controlled openness,’ that is more compatible with the identity of the EU and its policy approach to the external relations. To that aim, certain practical suggestions for the CJEU on how to implement that new paradigm inspired to the literature on comparative law are provided (Chapter V).
CHAPTER II – “AUTONOMY” OF THE EU LEGAL SYSTEM FOR THE CJEU
introduction
This chapter engages with the long-lasting debate among EU scholars on what ‘autonomy’ of the EU legal system is. It does so by developing a theoretical framework for the analysis of the content and use of autonomy of EU law on the part of the CJEU that will be relied upon for the remainder of this Ph.D. To that aim, section […] firstly sets the scene for this thesis by briefly looking at the characteristics and use of autonomy of legal persons in the context of legal philosophy. Section […] moves to compare the concept of autonomy in legal philosophy with the understanding of autonomy of the EU legal order by the CJEU. The section then focuses on two specific characteristics of the autonomy in the EU: its constitutional origin and how that origin affects the development of autonomy, making it a ‘dynamic’ concept. Specific relevance is given to the interface between autonomy of the EU with both the principle of effective judicial protection and dialogue with national and international courts. Having discussed the main elements of autonomy in the EU legal system as understood by the CJEU, section […] then moves to analyse its legal nature and develops a working definition for that concept that will be used for the remainder of this work. Finally, some brief conclusions follow in section […].
the theoretical background of autonomy
As Eeckes notes, etymologically, the term “autonomy” derives from auto (-self) and nomos (law) to mean “one who gives oneself one’s own law.” The term was first used in the work of Aristotle ‘The politics and the Constitution of Athens’ to describe the capacity of Greek city-States to shape preferences, create, and enforce law. The idea of ‘autonomy’ has in time been widely addressed in different realms – from education policy to biomedical ethics, to legal and freedoms rights – a long time before becoming relevant in the context of European Union law. Significantly, for Dworkin autonomy hardly has a general meaning and can only be understood in the specific context in which it is used. Along that line, he notes that “[A]utonomy is a term of art introduced by a theorist in an attempt to make sense of a tangled net of intuitions, conceptual and empirical issues, and normative claims.” A survey of the use of autonomy across different fields of studies indeed highlights that there is no predetermined list of what makes an entity ‘autonomous.’ Narrowing down the analysis to the field of legal studies and in relation to collective entities, relevant to this work, autonomy is normally regarded as the possession by an entity of a ‘distinct’ existence from other entities. Such distinct existence emerges from its: (i) self-determination (i.e., capacity to rule itself); and (ii) independence from external manipulations. These two elements are reflected in the concept of legal personality of collective entities, which essentially is the possession of a ‘separate will’ (political autonomy – independence condition), expressed through own institutions (institutional autonomy – self-determination), in accordance with the precepts of its constituent agreement. Self-determination and independence also evidence that autonomy intrinsically is a relational concept, in the sense of necessarily entailing some relationship with otherness. So, while some degree of independence and self-determination from the outside are still necessary for an entity to be able to determine its course of action, autonomy necessarily finds its roots in the existence of an external comparator. Autonomy also isn’t a one-dimensional concept, i.e., an entity can be autonomous to a certain degree. It follows from this that there are a number of possible shades of independence and self-determination of entities depending on their structure and functioning. It is that structure and functioning that determines their degree of autonomy from their constituents. That emerges from the classification made by Ege of the level of autonomy of twenty secretariats of international organizations, including the European Commission for the EU, for the purpose of ascertaining their potential influence on the decision-making of the international body. The author determines the degree of autonomy of the analysed secretariats based on their capacity to self-determine (defined as their ‘autonomy of will’) and implement their own preferences (defined as their ‘autonomy of action’), which, taken together, ultimately determine to what extent these secretariats are able to influence global public policy. Ege concludes his analysis by dividing the analysed secretariats in four categories (manager of status quo, politicized bureaucracy, ideational bureaucracy, and autonomous bureaucracy) depending on their degree of autonomy from politics within the organisation in which they are set up. Finally, it is also relevant that, as autonomy is shaped by the constituent agreement of each collective entity, it also normally is a dialogic concept whose content and characterisation can evolve in time. Constituent agreements that set up collective entities are living instruments, subject to the changing needs of the contracting parties via their interpretation or amendment, or to external pressures. Their provisions are often very broadly worded, and virtually all of them include amending tools. Therefore, autonomy of collective entities is bound to evolve as a result of the changes to the constituent agreement.
autonomy in the eu
Autonomy of the EU legal system as understood by the Court, shares many of the characteristics of the autonomy of collective entities discussed above. To begin with, analogously to what Dworkin notes in relation to the autonomy of legal entities, defining the ‘autonomy’ of the EU legal system is no easy task, and its understanding is highly dependent on the specific circumstances of each case. That is partly connected to the fact that the concept of ‘autonomy’ does not figure in the EU treaties, but rather is the product of the case law of the Court of Justice of the European Union (‘CJEU’ or the ‘Court’). The CJEU has consistently identified the autonomy of EU law with the fact that the EU cannot accept to: (i) be subject external interferences with “the essential character of the powers of the Community and its institutions as conceived in the Treaty,” and (ii) be bound to a specific interpretation of its rules adopted by international bodies. Yet, as for the concept of autonomy in legal theory, these guidelines of the Court are too broad to constitute meaningful guidelines to predict whether or not certain arrangements will in practice be considered compatible with the autonomy of EU law, and sometimes even to understand its judicial reasoning for excluding that compatibility. For instance, the ‘essential character of the powers of the Community’ has been regarded by the Court as encompassing a variety of different constitutional concepts and functioning mechanisms of the EU that are virtually able to describe most expressions of EU law, ranging from direct effect and primacy of EU law, to fundamental rights, from mutual trust, to common values and principles, and has been regarded as including the ‘community character of the law,’ the fundamental rules of the internal market and competition law, and the system of judicial protection of the EU, particularly the mechanism of preliminary reference under article 267 TFEU. Further, while the concept of ‘autonomy of EU law’ is also increasingly relied upon in acts of other EU Institutions in different contexts, no indications regarding its scope and meaning emerge from those documents. Secondly, the same two fundamental elements characterising autonomy in legal theory of ‘independence’ and ‘self-determination’ appear in the understanding of autonomy of EU law of the Court, in the form of the mentioned exclusion of interferences with “the essential character of the powers of the Community and its institutions as conceived in the Treaty,” and that the EU may be bound to a specific interpretation of its rules adopted by bodies external to it. Finally, as in legal theory, autonomy of EU law is also understood by the Court as a ‘relational’ and ‘dialogic’ concept. The one starting point on which all commentators seem to agree is that the Court identifies the roots of autonomy of the EU legal system in its claim of ‘constitutionality’ of the EU created in relation to other existing legal entities of national and international nature, particularly through a continuing dialogue with them. The next sections will analyse in more details these two characteristics of the autonomy of EU law.
3.1 Autonomy and the constitutional nature of the EU
The two features of the EU as an ‘autonomous’ entity and as possessing a ‘constitutional’ nature both emerge in the decision of the Court in Van Gend en Loos, where the EU was famously defined as a “new legal order of international law” formed of Institutions “endowed with sovereign rights” and able to impose rights and obligations on individuals “[i]ndependently of the legislation of Member States.” Reasoning that the mandate of the (at the time) Community was conferred upon the EU Institutions for the direct benefit of both the citizens and the Member States, the Court qualified the Union as a ‘new type’ of legal entity, characterised for being autonomous from both ‘national’ and ‘international’ law. The origin of autonomy in the constitutional claim of the EU is even more evident in the subsequent decision of the Court in Costa. The French version of the decision notably connects the nature of the treaty of Rome as “une source autonome” and “la base juridique de la Communauté elle-même.” To the knowledge of the author, Costa constitutes the first reference made by the Court to the ‘autonomy’ of the EU legal order. For the CJEU, like sovereign states, the EU is a homogenous legal system whose rules are applicable to all its participants: Barents’ idea of the “community character of Community law,” well expresses the concept of ‘unity’ and ‘indivisibility’ of EU law, which is intended to be (1) the same, (2) in all Member States, (3) in all circumstances. Like sovereign states the EU enjoys an interpretative monopoly on the effects of that law. Further, like for sovereign states, the law of the land creates immediately enforceable rights and obligations upon individuals – establishing an unseen direct link between the EU and the territories of Member States that international organisations normally lack. However, unlike sovereign states, the functioning of the EU is limited to certain competences and lacks one territory (instead having the ‘territories’ of the Member States), one population (only having the ‘peoples’ of Europe), and one government (being guided by Institutions with varying representation and composition). Therefore, the inherent right to self-determination of the ‘peoples’ of Europe and the sovereignty of the Member States can never be substantially absorbed in the EU due to its autonomous functioning alone. At the same time, the constitutional nature of the EU also made it different from an international organization for the CJEU, distinguishing the emerging new legal order from any other entity existing in international law. The ideological basis on which the Court carried out the ‘legal revolution’ that led to the ‘declaration of independence’ of the EU from its Member States and international law – through its ‘constitutionalization’ and ‘autonomization’ – in Van Gend en Loos and Costa, is the auspice for the “ever-closer union among the peoples of Europe.” That auspice, included in the funding treaties of the EU ever since the treaty of Rome in 1957, was arguably understood by the Court as the ‘ideological blueprint’ of the EU, a project common to all Member States, pursued for the benefit of the individuals: the creation of the new legal space of the EU aimed at the fostering of links and relationships among individuals of different nationalities, yet sharing a common ‘European’ background, to achieve a common ‘identity.’ the ‘ideological blueprint’ of the Union to come was, for the CJEU, the language used by the founding fathers of the EU: the term “peoples.” That term, which recurs in virtually all fundamental charters of sovereign states (albeit in its singular version of “people”), appears to have been understood by the Court as a unifying factor assimilated to the pouvoir constituent of nation states. In that context, the term ‘people’ does not refer to an aggregate of individuals but rather designs “persons who are part of a national community” as members of a political community, which share a common identity. The idea is that by bringing Institutions, Member States, and EU citizens under a specific set of rights and obligations the common law of the EU is able to transform states into ‘Member States’ and individuals into ‘citizens,’ as opposed to ‘third countries’ and ‘foreigners.’ This concept is well known in the field of political science of studies on democracy and identity, where it was noted that “what makes persons a people is simply this: co-existence, over time, under the rule of a given legal and political order. A people, for the purposes of democratic self-government, is the set of persons co-existing under the rule of a particular political order.” The Court however recognizes the essential difference in the use of the plural form of the term, ‘peoples,’ which it interprets as a rupture with the idea at the basis of nation states that common identities can only form within the borders of nation states. “Peoples” reflects the reality that the Union is not, nor it aims to becoming, a state. Quite the opposite. It wants to encompass multiple political communities on the territory of Europe sharing common ideological characteristics that the Union undertakes to protect and enhance through the construction of an economic community. It is in this ‘Copernican’ change of perspective from the traditional paradigm of international law of setting out a supranational system that recognizes individuals not only as subjects of law, but pivotal elements of a wholly new regime, that lies the fundamental importance of Van Gend en Loos for the constitutional structure and autonomy of the EU. Reasoning on the willingness of Member States to limit their own sovereignty to confer sovereign rights to a Europe which brings together governments to create one common European identity for the benefit of the individuals, the Court qualified individuals as ‘subjects’ of (at the time) EC law, creating a direct link between the EU and its peoples. That central role of individuals in the EU – conferred by the Court the until then unseen role of “cooperat[ing] in the functioning of the Community” – also allowed the Court to ensure the effectiveness of the new-born legal system of the EU. As Mancini notes, Van Gen den Loos further empowered individuals by making them essential enforcers of the EU legal system through EU-specific mechanisms of protection. Through the ‘European’ courts, i.e. the national courts of the Member States under the supervision and scrutiny of the CJEU, Van Gend en Loos sets the basis for individuals to exercise their EU-law derived rights independently from – and even in opposition to – their national governments, making these latter accountable before the Court and ensuring the capillary penetration of EU law across the legal system of the Member States.
It will be seen later in this work that that feature of the EU, which will be subsequently developed by the Court into a full primary law right to ‘effective judicial protection,’ has in the years acquired a progressively more important function in its constitutional framework: from a general tool of enforcement of the law, to an instrument to curb differences across legal procedures in the Member States, further to an instrument to require their active compliance with principles of EU law, to become today a constituting element of the identity of the EU able to impose a common EU-specific justiciable threshold of protection of the rule of law across the legal systems of 27 Member States. It thus does not come as a surprise that, as it will be discussed in more detail later in this work, the Court regards the principle of effective judicial protection as a key element of the constitutional identity of the EU, one that cannot be compromised by outer interferences on pain of compromising the autonomy itself of the EU. It follows from the above that autonomy is a founding element of European constitutionalism and an affirmation of the identity of the new-born legal system crafted around the creation of the ‘European peoples’ and of its distinguishing characteristics, particularly the centrality of individuals and their rights through their judicial protection.
3.2 Autonomy and the dialogical nature of the EU
The ideological aim of creating the “peoples of Europe” was however still insufficient to support the ‘declaration of independence’ of the EU. The case law of the CJEU seems to support the idea that the ‘ideological mandate’ of the EU is associated with a second, ‘voluntarist,’ element of individuals and Member States to be involved in a shared project of unification under a common, European, set of liberal values and principles. This voluntarist’ element forms an integral tile of the constitutional nature of the EU and legitimises its autonomy from national and international law. For the purpose of this Ph.D., it is interesting to note that the courts and tribunals – national, European, and international – played a pivotal role in the construction of the constitutional project of the EU by shaping what the EU is and legitimising its existence vis-a-vis both national and international law. A crucial role in this process was played by the CJEU, in its double role as the ‘domestic’ and ‘international’ court of the EU. Firstly, in the internal relations with the national judiciaries, as a Supreme Court (highest judicial tribunal in the political unit of the EU), Court of Cassation (jurisdiction to resolve interpretative questions about the relevant law in a uniform manner), and Constitutional Court (jurisdiction to ensure that the spirit and norms of the EU treaties are upheld) of the EU, the CJEU has played a key unifying and legitimising role by centralising litigation in the EU through EU courts. Through dialogue with national courts, the Court firstly involved individuals and Member States in the creation of a common EU identity, by adding to the few explicit constitutional principles of the founding treaties of the EU new elements to create a European identity. These famously include the distinctive features of the EU of principles of primacy, direct effect, effective judicial protection, fundamental general principles of law, and even fundamental rights. The internal judicial system led by the CJEU also contributed to cross-fertilization into the legal system of the Member States, curbing differences across borders and promoting unity through the creation of common principles, common standards, and common values. Through litigation the Court was further able to ‘test’ the position of the Member States and Institutions on the status of the law in a view to put forward new, at times bold, claims regarding the constitutional nature of the EU, starting an intersubjective discourse with outer entities, national courts, individuals, Member States – that constitutes the motor of the development of the EU. The CJEU also played the pivotal role of creating internal legitimacy for the EU legal system by acting as ‘sounding boards’ for the endorsement by Member States and individuals of the Court’s decisions and their implementation in the form of all, treaty amendments, compliance with EU law, and litigation before EU Courts: through this prism of voluntarism the legitimacy of the EU legal system is pictured as an ongoing process in which participate both individuals (for the creation of whose European political identity the EU exists) and Member States (masters of the treaties and addressees of Union’s legislation) as “interested parties of the Community.” The more Member States accept to be bound by the scheme of the EU treaties, the more they exercise their rights and fulfil their obligations under EU law, the louder their consent to being part of the EU legal system. Equally, the more individuals rely on their rights under EU law, the more they unhinge the traditional state structures by rendering accountable their Governments before the EU Institutions, the more they pressure their home states to participate in the unification of the EU legal system for the achievement of the European project. Secondly, in the external relations, wearing the double hat of international court and constitutional court of the EU, the CJEU has acted as a catalyst of dialogue between the EU and international law shaping the identity and building the legitimacy of its legal system vis-à-vis international law. Dialogue with international law through the CJEU undoubtedly contributes to shape what the EU is – and what it aims to become – in relation to foreign actors and legal systems. One may use the metaphor of international law as a ‘mirror’, which, through the mediation of the CJEU, progressively contributes to shed light on the values and priorities of the EU compared to others. For instance, in time the catalyst role of the CJEU has defined the role of international law in the EU treaties, borrowed and readapted rules of international law to the EU legal system, defined whether and to what extent decisions of international courts can be implemented in the EU, as well as the essential tenets of the constitutional legal order that can never be compromised in the relationship with international law. For instance, in decisions such as Opinion 2/13 the Court was able to clarify that the EU can only be subject to international law for as long as such law complies with its constitutional principles, including rule of law, judicial protection, human rights, and democracy. Similarly, on the topic of this Ph.D., the Court has questioned the capacity of international dispute settlement mechanisms to deliver effective judicial protection to a comparable standard to the one existing EU-wide. The external dimension of the judicial dialogue of the CJEU also contributed to legitimising the developing EU legal system in the eyes of international law. While the international legal personality of the EU was not always self-evident in the face of the EU treaties, participation of the EU, also through its Court, in international law contributed to its recognition as a fully-fledged legal actor. Decisions in which international courts take account of the existence and development of the EU legal system, legitimise and promote the Union’s international actorness and make its voice hear in international law. This is true both for those decisions that endorse certain characteristics of the EU legal system, such as the Bosphorus presumption of the equivalence of the protection of ECHR rights by the EU; and for those decisions that reject them, such as the numerous decisions of investment tribunals rejecting that decisions on their jurisdiction on investment disputes may be decided under EU law. Notably, the decisions of the CJEU have also influenced other Courts internationally, contributing to export the ‘EU rule of law’ abroad. As it will be seen in more details below, not only these decisions played a role in the interpretation of principles applicable in different jurisdictions, but they also impact on the creation and functioning of international mechanisms – as it was the case for the system of listing of sanctions on the part of the UN Security Council (UNSC), the creation of an international investment court, or the project of legal unification around the globe, through the creation of no less than 13 ‘copies’ of the judicial system of the EU, which have explicitly embraced the principles of supremacy and direct effect doctrines developed by the CJEU. It follows from this that, as Lenaerts also observes, the unique constitutional nature of the EU and its autonomy were never meant to be complete isolation from international law, but find their root in the ‘esprit d’ouverture’ that inspired the authors of the EU treaties after the ravages of the second World War and a balancing exercise between the construction and safeguard of a constitutional space in Europe and interaction with the wider world.
In conclusion, the voluntarist element of dialogue with national and international law has arguably accompanied the development of the identity of the EU from a hybrid national-international law entity into a fully-fledged constitutional actor, consolidating its structure, enlarging its competence, defining its boundaries, and strengthening its powers vis-à-vis national and international legal orders. As Alter points out, few would question today some established principles of EU law like primacy and direct effect. The EU treaties have undergone a number of amendments in time and, despite the failure of the constitutional treaty, today include all the elements of a traditional democratic constitution: values, hierarchy of sources, fundamental rights, legislative procedures, and even tools to ensure a state-like ‘distribution of powers.’ The legal capacity of the EU, both domestic and international, is now a fact. With occasional hiccups, national courts also hardly question their position as first point of delivery of EU law and take seriously their essential role in the iterative process of development of EU law through the preliminary reference procedure in article 267 TFEU. It thus seems reasonable to conclude that the constitutional claim of the EU, its autonomy, and even its identity, are based on pluralism and exchanges with national and international law, as a natural consequence of a decentralised legal order with a specific history and own aims. Autonomy of the EU legal system is inevitably a part of the dialogical development of its constitutional structure achieved through continuous exchanges with domestic and international law. Thus, analogously to what was seen for the concept of autonomy in legal philosophy, its content can never be understood as static, but it physiologically develops together with the evolution of the constitutional identity of the EU.
4.autonomy as a doctrine of constitutional interactions
As Odermatt notes, while the Court does not directly engage in discussions about its nature, autonomy has been qualified by commentators in multiple ways – as a principle, an idea, or a doctrine. Given the different functions of each one of these legal concepts – respectively, a guideline for interpretation, an amorphous non-legal concept, and a fundamental proposition about the law by the courts – the qualification of autonomy as falling in one or another category is bound to have significant consequences in terms of the understanding of its effective role and practical effects on the EU legal systems. As a preliminary point, the specifically legal significance that the Court attaches to autonomy as an element of the constitutional structure of the EU, excludes that autonomy may be regarded as an ‘idea.’ Similarly, its blurred scope and non-prescriptive content exclude that autonomy may be regarded as a ‘rule,’ defined by Tiller and Cross as a “strict requirement that define the answer to a dispute, once the predicate facts are established.” The issue remains of whether autonomy is to be regarded as a ‘general principle,’ which orients the interpretation of EU law, or as a ‘doctrine,’ which shapes the identity itself of the EU and is able to orient the creation of new law. The sections below analyse the main characterising traits of doctrines, as opposed to principles, and compare the resulting considerations to the characteristics of autonomy of the EU legal system as discussed until now. This thesis supports the argument that autonomy is as a ‘doctrine’ of EU law for three main reasons.
4.1 Autonomy is not a principle because it is not a source of law
A preliminary observation is that doctrines, especially in common law countries, are mostly created by courts, while in civil law countries they also encompass authoritative scholarly writings. Considering that in the EU both general principles and ‘autonomy’ find their origin in the case law of the Court, that consideration does not constitute a determinative clue on the nature of autonomy. There however appears to be some agreement among scholars that doctrines are not a source of law. One of the earliest theorists of the theory of legal doctrine, Alf Ross, only regarded as sources of law legislation, custom and precedent, and reason. Ross argued that doctrines cannot instead be considered as sources of law, but rather they only provide ‘valuable support to practice.’ A similar position is shared by Planiol and Ripert, who observe that even unanimous doctrines cannot constitute a source of law. Atwill also concludes that “a doctrine is not a recognized source of law as it is not vested with any institutional power that authorizes it to create law.” Doctrines are rather considered by the author as a de facto authority. General principles are considered supplementary sources of EU law. The fact that autonomy is has not, to the knowledge of this author, been used by the CJEU as a source of law, but only as the ‘general constitutional backdrop’ for its decisions (see below), would appear to be a first argument in support of the thesis that autonomy is not a general principle of EU law.
4.2 Doctrines as courts’ theorization of what the law is.
Many authors seem to agree that the first main function of doctrines is to explain all the existing rules of law. Unlike principles, which guide the application and interpretation of the law towards what the law should be, doctrines constitute the understanding of the Court of the fundamental backdrop of the existence of the law at the present state, based on social, as much as legal, considerations. Ross, for instance, defines doctrines as “cognitive assertions concerning valid law of greater or lesser degree of probability.” The author regards them as the reflection of social events, impossible to verify objectively. Tiller and Cross regard doctrines as “the currency of the law […] doctrine, or precedent, is the law, at least as it comes from courts” which “sets the terms for future resolution of cases in an area.” A similar approach is adopted by the literature in the field of European law: Lasok and Bridge regard doctrines of EU law as a “general proposition” or “guidance relating to a fundamental issue such as eg the nature of Community law.” The authors note that fundamental doctrines of EU law include, for instance, primacy and unity of EU law. Lasok and Bridge also note that doctrines as systemic assumptions have a too broad scope to be applied in themselves to a certain set of facts or sanctioned. The authors also identify an essential distinguishing element of doctrines from general principles of law in the impossibility to directly apply doctrines to specific cases, which they instead view as working as ‘rules of conduct.’ These characteristics of doctrines align with the understanding of the CJEU of the constitutional nature and functions of autonomy of EU law. To begin with, in line with the literature cited above, autonomy normally represents the view of the Court of what EU law is, rather than the application of what it should be. As an essential element of the constitutional identity of the EU, autonomy describes what the CJEU understands the main characteristics of the EU legal order to be – including values, rights, and general principles, as well as the mechanisms necessary to implement them – which form the essential tenets of the identity of the EU. Further, the understanding by the CJEU of autonomy possesses the same ‘meta-legal connotation’ that Ross attributes to doctrines, as it encompasses reflections of social events – i.e., the ideological blueprint of the creation of a European identity – which cannot objectively be verified. The meta-legal connotation of autonomy is a consequence of its constitutional origin and dynamicity discussed above, on the basis of which autonomy evolves together with the identity of the EU. That is for instance evident in relation to fundamental rights. Fundamental rights, relied upon by the Court in Kadi to limit the impact of international law over EU law on the grounds of autonomy, were indeed introduced by the Court itself in Nold and Internationale on the basis of the unverifiable assumption that they are an integral part of the constitutional identity of the EU. That statement was at the time to be found nowhere in the EU treaties, so that the characterisation of fundamental rights as constitutional traits of EU law constituted more a leap of faith than a strictly legal reasoning. In addition to this, and in line with the account of ‘doctrine’ of Lasok and Bridge, unlike general principles, which are typically relied upon to the specific case at issue as ‘aids to interpretation,’ ‘gap-fillers,’ and ‘grounds for judicial review’ – as it was the case, for instance, respectively for the principle of non-discrimination in Mangold, the principle of institutional balance in Les Verts, or the principle of legitimate expectations in in the Staff Salaries case – doctrines cannot be ‘applied’ to a specific set of facts. They rather systematise the facts of a case and form the theoretical backdrop on the basis of which principles will then be applied to solve a specific case. From this standpoint, doctrines may perhaps be better understood as forming the theoretical background against which general principles are developed: so, to use a metaphor with a painting, doctrines may be compared to a canvas – the fundamental basis in the absence of which painting cannot exist –, general principles are the frame within which the artwork is developed, while rules and policies are the brushstrokes. The case law of the Court arguably supports the suggested framing of autonomy as a doctrine. To begin with, the Court assesses the compatibility of national and international law with EU law in relation to principles and mechanisms that constitute an expression of the autonomy of the legal system – rather than directly in relation to ‘autonomy’ in itself. For instance, in the case of Opinion 2/13 the Court qualified the judicial system of the EU as functional “to ensure that the specific characteristics and the autonomy of that legal order are preserved.” Similarly, in Achmea the Court reasoned that the purpose of the exclusive jurisdiction of the Court under article 344 TFEU is a tool “to safeguard the allocation of powers laid down by the Treaties, and hence the autonomy of the EU legal system.” Using the language of the Court itself, autonomy functions as the background to the tools and principles that form the “fundamental premiss” of the relationship among EU legal system and other systems of law, i.e., the essential canvas on the basis of which the general principles of EU law – such as fundamental rights –, and the essential elements of the functioning of the EU legal system – such as the system of judicial protection of the EU (267 and 344 TFEU) – operate.
4.4 Doctrines as influencing future legislation and case law
From another standpoint, authors concur that the second main function of doctrines lays in their ability “to influence the legislator in framing new legislation and guides judges in their application of existing laws.” The point highlights the specific characteristic of doctrines of sitting at the intersection between politics and law, which emerges from the same etymology of the word: ‘doctrine’ derives from the ancient Greek word ‘δόξα,’ which originally means ‘common opinion’ or ‘common belief’, as opposed to ‘ἐπιστήμη,’ which defines ‘knowledge.’ Numerous authors concur with this reconstruction of the function of doctrines. As Ross also notes, however theoretically aimed at defining what the ‘likely’ legal developments, doctrines remain a “legal-political” factor, influenced by an underlying ideology and itself influencing the future of that development. Similarly, studies on the development of courts have highlighted that ideology can explain why judges choose a path over another and influence, through the strong legitimating function of adherence to a specific doctrine, subsequent decision-making. Similarly, Zahle notes that legal doctrines take account of the state of the law and make predictions of how it will evolve in the future, in so influencing that development. Zahle further observes that the reactions to the doctrine may accept the statement of the court or falsify it, and, in that case, they reflect on the validity of the doctrine. Therefore, doctrines are bound to be heavily relied upon in particularly problematic cases, where they lack precise guidance, and can be used to reframe or push the boundaries of what the law is through the expansion of precedent. This “political-legal” nature of doctrines is not shared by general principles, which rather serve the aim of influencing the application of the law, rather than its creation. The nature and development of the autonomy of the EU legal system in the case law of the CJEU seems to once again support its characterisation as a doctrine rather than a general principle. To begin with, it was noted above in section […] that autonomy is understood by the Court as possessing a strong ‘ideological’ component (referred to above as the ‘ideological blueprint’ of the Union), which orients the development of the legal system towards constitutionalization. The original ideological blueprint of the “ever closer union” for the benefit of the peoples of Europe recurs in the case law of the Court on autonomy almost as a ‘self-fulfilling prophecy’: through autonomy the CJEU ‘discovers’ a not self-evident political and ideological nature of the EU. That political and ideological nature of the EU after Van Gend en Loos becomes an assumption in its case law, first, and, increasingly, in the case law of the national courts of the Member States and in the political and legal debate surrounding the nature of the EU. The nature of doctrines as “legal-political” factors able to influence legislators and guide judges also fits well with the ‘dialogical’ nature of the EU legal system discussed in the previous section: through the elaboration of the doctrine of autonomy, the Court ‘tests’ the limits of what national courts and Member States may be prepared to accept on the grounds of the ‘constitutional’ nature of the EU and assesses whether to ‘push the boundaries’ of the law. As Lazowski points out, in Van Gend en Loos the Court “gambled on its own legitimacy. When declaring that the then Article 12 of the EEC Treaty was capable of producing direct effect, the Court of Justice exposed itself to criticism and a potential backlash from the Member States.” The CJEU continued to ‘gamble’ in the years to follow. National courts and Member States have often accepted the interpretation of the Court of what the implications are of the constitutional nature of the EU are, as they have done in Van Gend en Loos as much as, for instance accepting principles which find their source in creative decisions of the Court, such as the international legal personality of the EU or the protection of fundamental rights. Most of these ‘gambles’ of the Court have notably been formally constitutionalised, through their inclusion in the EU treaties, but not only. It is relevant for the purpose of this Ph.D. that views of the Court on autonomy of the EU from outer scrutiny have in the past heavily influenced the negotiations of the dispute settlement mechanisms set out in many international agreements of the EU. That was for instance the case of the Draft Agreement to the European Convention on Human Rights, or the Investment Court System (ICS) in the trade agreement between the EU and Canada, which were provided with mechanisms developed around the case law of the CJEU on autonomy. Yet, as for all doctrines, national courts and Member States may also ‘falsify’ the Court’s interpretation of the ‘implications’ of the constitutional nature of the EU, as they have done by, for instance, excluding that the EU treaties could formally become ‘constitutional charters’ of the EU. Finally, one may also see a confirmation of the nature of autonomy as a ‘doctrine’ of EU law in the fact that the Court only uses the term ‘autonomy’ in particularly complex cases where it most evidently introduces ‘daring’ interpretations of the EU legal system, which almost require a ‘leap of faith’ on the part of the recipients of its decisions. That is for instance the case of Opinion 2/13, or Opinion 1/17, which will be analysed below in more details.
4.5 Autonomy as a doctrine ‘of constitutional interactions’ and its relevance
The sections above support the claim that autonomy is understood by the CJEU as a doctrine: it is not in itself a source of law; it constitutes the interpretation of the judiciary, the Court, of what the law is; and influences the creation of new law on the part of the legislator. It follows from this that, firstly, as a product of the constitutional nature of the EU, autonomy needs to develop within the constitutional framework of the treaties and cannot unilaterally modify them. Indeed, once the CJEU provides one possible explanation of what the law means, the other participants in the EU legal system may reject that explanation by ‘falsifying’ it. Secondly, as autonomy entails broad assumptions on how the law will develop, it normally confers to the CJEU large discretion in the interpretation of EU law – even greater than the one it possesses in the application of other broad concepts of EU law, such as general principles – and forms the springboard through which the Court can influence the development of the legal system. To the claim that autonomy is understood by the Court as a ‘doctrine,’ it can be added that it is also understood as having as its subject matter ‘constitutional interactions.’ On the one hand, the reference ‘constitutional’ is meant to highlight that the CJEU uses the doctrine of autonomy in relation to matters affecting the identity itself of the EU, such as its constitutional nature or its basic functioning mechanisms. On the other hand, the term ‘interactions’ evokes the dialogic nature of the EU, which, as discussed in chapter […], is not fixed in time but develops through a constant flux of interactions among the participants in the legal order. The term ‘interactions’ also wants to evidence the relational element of autonomy discussed in chapter […] with regards to the significance of autonomy in the broad field of legal studies: autonomy entails the existence of some relation between the self and others (autonomous from what?) and is at the odds with autarchy. ‘Interactions’ thus intends to transmit the idea that the EU develops through dialogue and comparison with others. On the basis of the above, it is now possible to develop a working definition of autonomy. For the purpose of this work, autonomy as ‘a doctrine of constitutional interactions’ is ultimately understood in this work as a judge-made concept that defines the conditions and content – progressively codified in the constituent agreement of the EU, the EU treaties – that the constitutional identity of the EU sets to the interactions between the EU and entities sitting outside of its legal order. In terms of its broader relevance to the current debate on autonomy, this definition differs from the discussions on the topic to date in so that it broadens its scope and relevance in the case law of the Court. As the fundamental element of the identity of the EU, autonomy constitutes the baseline of all the decisions in which the Court has engaged with the constitutional interactions of the EU and national and international law. So, it will be argued in chapter […] below, the discussion on autonomy of the EU from international law cannot be limited to that handful of decisions in which the Court expressly mention the word ‘autonomy.’
Is autonomy a ‘structural principle of external relations law’?
Separate mention deserves the theory developed by Marise Cremona of autonomy as a ‘structural principle of external relations law’ of the EU. Various reasons justify the specific relevance of that theory for this Ph.D. and its specific analysis and comparison with the definition of autonomy as a doctrine of constitutional interactions. These are not only the authoritativeness of its author, but also its timing – the theory was only recently developed –, as well as the fact that the ‘structural principles’ discussed by Cremona can hardly be framed in the existing classifications of law discussed above of general principles of law, concepts, or doctrines, which was used as a baseline for the analysis. Furthermore, the theory of structural principles developed by the Cremona notably focuses on autonomy specifically from the standpoint taken by this Ph.D. of the external relations of the EU, and how that principle influences the Union’s relationship with international law and outer scrutiny. Cremona defines ‘structural principles’ of EU law as a “set of inter-locking principles” which define the “framework expressed in (or implied from) the Treaties, protecting an institutional space within which policy may be formed, in which the different actors understand and work within their respective roles.” The argument put forward by Cremona is that its reticence to interfere with policymaking in the external relations has lead the CJEU to reframe its influence through the definition of “the type of international actor that the EU is.” That definition would develop by norms framing the processes in the external relations of the EU – as well as the relationships between the actors involved in them – through which the EU interacts with international law. Cremona defines this new category of norms as ‘structural’ in the sense that “they help to define the structure of the EU as an international actor, both internally and externally.” The author explains that these norms act as ‘principles’ because of their legal nature as fundamental rules of EU law, their general content, and the fact that their breach “may result in the illegality of the resulting measure.” Cremona also separates the concept of ‘structural principles’ from the traditional category of the ‘general principles of law,’ in so that they regard the form of the policy-making activity of the EU rather than the substance of EU law. The author admits however the existence of some overlaps between the two categories as some of her ‘structural’ principles – such as effectiveness, transparency, and proportionality – also are general principles. Structural principles are further defined by the Cremona as pertaining to the sphere of ‘external relations law’ – i.e., the relationship between the EU and international law – because, although being relevant for the relationship of the EU with both its Member States and international law, in international law they are not so much concerned with the object of EU aims (e.g., market access) but rather with the subjective identity of the EU as an effective policy-maker. Cremona further divides structural principles in two sub-categories: ‘relational’ and ‘systemic’ principles. ‘Relational’ principles, such as the principle of conferral of powers, common values, and mutual trust, “govern the relationships between actors or legal subjects (not norms)” and “structure the framework within which the actors in the EU’s system of external relations can play their roles, deciding and implementing policy.” ‘Systemic’ principles, which include autonomy, are instead viewed as “concerned with the operation of the system as a whole, with building the EU’s identity as a coherent, effective and autonomous actor in the world.” Unlike relational principles, systemic principles not only frame, but also guide, the dynamic development of the EU as a whole, in a manner that responds to the principles enshrined in the EU treaties. It is argued that the working definition of autonomy developed in this Ph.D. as a ‘doctrine of constitutional interactions’ of the EU partially overlaps with the insightful reconstruction of autonomy as a ‘systemic structural principle’ of the EU described by Cremona. Firstly, as mentioned above, the qualification of autonomy developed in the present work encompasses, among other things, the ‘structure’ of EU law: autonomy was indeed regarded above as a ‘fundamental canvass’ on which the EU bases its relations, both with internal and external actors sitting outside of its legal system. Thus, analogously to Cremona’s reconstruction, the definition of autonomy as a ‘doctrine’ also is inherent to the deep structure, and the identity itself, of the EU. Further, this work agrees with the insightful analysis carried out by Cremona that autonomy is constitutional in nature, as it is rooted in the EU treaties and was further elaborated by the Court based on that constitutional nature. Moreover, it will be seen below in chapter […] that, analogously to the analysis of Cremona, autonomy as a ‘doctrine of constitutional interactions’ also covers processes that, however different, belong to the relations of the EU with both national and international law.
However, four fundamental differences exist between the working definition of autonomy as a ‘doctrine of constitutional interactions’ in this work and the definition put forward by Cremona of autonomy as a ‘structural principle of EU law.’ Firstly, the different nature and functions of principles of law compared to doctrines discussed above. Secondly, Cremona appears to view autonomy as purely structural, as a limitation put by the Court to the discretion of policymakers in the external relations. On the opposite, the definition of autonomy put forward in this work as ‘doctrine of constitutional interactions’ also entails a substantive feature and has a positive, in addition to negative, side: setting out the ‘content and conditions’ of the Union’s interactions with other entities, autonomy defines the identity of the EU in terms of what the EU is and can do, and not merely the its structural limits, i.e., what it is not and it cannot do. As an essential element of European constitutionalism, autonomy is instead understood here as being part of the identity of the EU, encompassing procedures as much as substantive values. It follows from this that autonomy as understood by the Court not only sets the limits to the interactions between the EU and other national and international entities sitting outside of its legal system. Autonomy itself allows and is part of those interactions. A further consequence of this definition of autonomy is the rejection of the very insightful distinction made by Cremona of ‘structural principles of external relations’ in ‘relational’ and ‘systemic’ principles, where autonomy would fall within that second category. That distinction arguably has little significance from a standpoint of autonomy elaborated here: if autonomy is regarded as encompassing the identity traits of the EU on the basis of which the EU develops its internal and external relations, then it presides to both sets of relationships. Even more fundamentally, Cremona views autonomy as an attempt by the Court not to impinge in the policy discretion of the political institutions of the EU. The definition of autonomy as a doctrine entails exactly the opposite. It was seen above that, hrough the use of autonomy as a doctrine, the Court carves for itself a very large space for decisional discretion in the external relations. A discretion which allows it to adopt an ‘activist’ role in relation to the international activity of the Union by reserving for itself an important position in the definition of the international actorness of the EU and pushing the boundaries of that definition. For all these reasons, despite the large number of overlaps, there appears to be some irreconcilable differences between the working definition of autonomy adopted in this work as a ‘doctrine of constitutional interactions’ and Cremona’s ‘structural principles of external relations law.’
Conclusions
This chapter attempted to contribute to the existing literature on autonomy of EU law by elaborating a theoretical framework for its nature and use by the Court. Section […] sets the scene for this thesis by giving an overview of the main elements and components of autonomy in legal theory relating to collective entities, identified in self-determination and independence from external influences. Autonomy was also regarded as a relational, multi-dimensional, and dialogic concept. Section […] drew some parallels between the concept of autonomy of collective entities and the autonomy of EU law as understood by the CJEU. It then focussed on two specific features of autonomy in the EU legal system: its constitutional origin and its dialogic nature. Building on the existing doctrine, the section argued that autonomy represents a founding element of European constitutionalism, an affirmation of the identity of the new-born legal system crafted for the benefit of the ‘European peoples’ and of its distinguishing characteristics. Among the distinguishing characteristics of the EU, specific importance possesses the protection of individuals and their rights, well-expressed in the right to judicial protection. Autonomy is also a physiologically dynamic concept, which naturally develops through dialogue with national and international law. That dialogic relationship is an essential motor for the evolution of the constitutional identity of the EU and its legitimacy vis-à-vis national and international law. Finally, section […] set out a working definition of autonomy as understood by the Court as a ‘doctrine of constitutional interactions’ between the EU and outer entities. Drawing a difference with general principles, the section argued that, as a doctrine, autonomy: (a) is not in itself a source of law; (b) constitutes the interpretation of the judiciary, the Court, of what the law is; and (c) influences the creation of new law on the part of the legislator. For this reason, it was argued, the doctrine of autonomy reflects an understanding of the Court of the traits of the identity of the EU and functions as the fundamental canvass on which the legal system is construed. However, while autonomy as a ‘doctrine’ confers a very broad discretion onto the Court for the interpretation of the conditions and limits to the interactions with international law set by the identity of the EU, it must be developed within the constitutional framework of the EU treaties. Despite the term being expressly used in only a handful of particularly controversial decisions, autonomy underlies all the constitutional interactions between the EU and international law. The next chapter will now examine the main traits of the constitutional and policy framework presiding to the relationship between EU law and outer scrutiny of international courts and tribunals, with a specific focus on arbitration and investment arbitration.
CHAPTER III – THE EVOLVING ROLE OF INTERNATIONAL LAW IN THE TREATIES AND EXTERNAL AUTONOMY
1. introduction
Chapter II argued that the CJEU uses ‘autonomy of EU law’ as a ‘doctrine of constitutional interactions’ between the EU and entities sitting outside of its legal system. Yet, while autonomy remains a ‘view’ of the Court, that view necessarily need to develop within the legal framework of the EU treaties and in dialogue with national and international law. This chapter moves on to analyse the development of the constitutional and policy framework presiding over the interactions between the EU and dispute settlement mechanisms carrying out outer scrutiny over the EU, and how that framework impacts on the interpretation of the doctrine of autonomy on the part of the Court. The chapter is structured as follows. Section […] introduces a distinction between two ‘sides’ of autonomy, internal and external, and clarifies that this Ph.D. focuses on external autonomy. Section […] argues that the Court originally adopted an approach of ‘separation’ from both national and international law in a key to support the construction of the EU constitutional space against external interferences with its fundamental characteristics, particularly the principle of judicial protection. That approach was supported by the ‘residual’ role of international law in the EU treaties of the origin which arguably also took stance of the greatly more advanced development of the EU legal order and its system of judicial protection over any other international dispute settlement mechanism. Section […] evidences a shift in the role and functioning of international dispute settlement mechanisms in international law: a ‘second generation’ of courts and tribunals is ‘delegated’ by states and international organisations vast authority to help enforcing the rule of law internationally. A delegation that resembles, and sometimes even goes beyond, the one conferred to the CJEU by the EU. While engagement with these dispute settlement mechanisms requires giving up some sovereignty, their nature as powerful tools of global governance, presents multiple incentives for the EU to engage with them. Section […] evidences that the EU executive has increasingly accepted the compromise underlying that by adopting an approach of progressive more openness towards these mechanisms. A similar trend is found in the EU treaties: section […] evidences how an approach of dialogue and cooperation with international law and international dispute settlement mechanisms over one of separation has become increasingly central not only to the achievement of the objectives of the EU treaties, but to its own identity. Section […] concludes that the principled approach of cooperation and dialogue with international dispute resolution mechanisms found in the EU treaties and embraced by the EU executive should therefore also be reflected in the understanding of the Court of external autonomy from outer scrutiny. Some brief conclusions follow in section […].
2. two sides of autonomy: internal and external autonomy
Authors working on the ‘autonomy of EU law’ seem to agree that, analogously to collective entities, the autonomy of the EU legal order system possesses two main dimensions, ‘internal’ and ‘external.’ Based on the working definition of autonomy elaborated in the previous chapter, ‘internal autonomy’ can be defined as a judge-made concept that defines the conditions and content – progressively codified in the constituent agreement of the EU, the EU treaties – that the constitutional identity of the EU sets to the interactions between the EU and its Member States. ‘External autonomy’ also is a product of the case law of the CJEU, which however defines the conditions and content – progressively codified in the constituent agreement of the EU, the EU treaties – that the constitutional identity of the EU sets to the interactions between the EU and international law. This work focuses on this second dimension of autonomy. Given the breath of all potential ‘constitutional interactions’ of the EU with international law, the Ph.D. focuses on the understanding and application of ‘external autonomy of EU law’ of the CJEU in relation to the outer scrutiny of dispute settlement mechanisms, taking as a case study arbitral tribunals and investment tribunals. As Koutrakos notes, it is often difficult to draw a line between internal and external autonomy, as most of the constitutional interactions of the EU with international law are, more or less evidently, a reflection of its internal constitutional structure. The author refers, for instance, to the internal and external constitutional implications of the Kadi decision for the EU legal order, where the Court found the unlawfulness of a EU regulation transposing a UNSC Resolution on the ground that the procedure on which the resolution had been adopted violated the right to effective judicial review and the right to property of the claimants: on the one hand, the judgement intended to safeguard the integrity of the system of fundamental rights of the EU from outer interferences. On the other hand, it also wanted to protect the relationship of the CJEU with the national courts of the Member States, by preventing the emergence of unwelcome Solange-type dynamics. The different use that the Court makes of the doctrine of autonomy in the external and internal relations of the EU raises interesting questions in relation to its potentially different role and rationale in the two dimensions. The CJEU initially elaborated the doctrine of autonomy primarily for the internal purpose of creating a constitutional space for the development of a new system of law. It was seen above that the first reference to that doctrine was found above in Costa, a decision with markedly internal importance, i.e., the establishment of primacy of EU law over the law of the Member States. That is understandable, considering that the CJEU based the development of the constitutional nature of the EU legal system primarily on those internal interactions. Yet, a survey of the database of decisions of the Court of Justice of the European Union shows that the term ‘autonomy’ is normally used by the Court to discuss the content and conditions of ‘constitutional interactions’ in the field of the external relations, particularly in relation to outer scrutiny of international dispute settlement mechanisms. In particular, after Costa the doctrine re-appears in 1976 in the context of an opinion on the compatibility of international agreements with EU law under article 218(11) TFEU and, later on, in a handful of judgements, mostly on the relationship between EU and international law. There seems to be no indications in the case law of the CJEU of why, after Costa, the Court used the doctrine of autonomy of EU law mostly in the context of the external constitutional relations. The constitutional relations between the EU and its constituents are arguably at the root of the EU legal system and certainly no less important than the ones with international law. One may speculate that the rationale of that omission is that the doctrine of autonomy of the EU legal order is extremely politically charged: it was seen above that autonomy, as used by the Court as a doctrine of EU law, entails more than the purely legal agreements set out in the EU treaties. It instead takes into account social considerations and sets out the idea of the Court of what the law is, in so heavily influencing its future development. For these reasons, the doctrine of autonomy opens the door to speculations on the significance and scope of the common European project, which remains contentious among Member States. Thus, the decision to only resort to autonomy in the framework of the constitutional relationships of the EU with international law may well simply constitute the outcome of a careful choice of wording on the part of the Court rather than a signal that any meaningful differences exist in relation to the concept of internal and external autonomy. Another explanation for the lack of use of the doctrine of autonomy of EU law in the internal relations is that, in that context, the Court arguably relies on another pillar doctrine of EU law to achieve the same results of delimitating the features of the actorness of the EU, i.e., ‘primacy.’ That was evidenced since Van Gend en Loos where the Court noted that the constitutional nature of the (at the time) Community was the outcome of a willing choice, “for benefit of which the states have limited their sovereign rights.” The all-encompassing function of the doctrine of supremacy thus arguably leaves no real use for the application of autonomy in the internal relations.
3.the approach of the origins: autonomy as a tool of separation
3.1 ‘Construction and safeguard’ and the development of the judicial protection
In the discussed balancing exercise between the ‘construction and safeguard’ of a constitutional space in Europe and ‘interaction’ with the wider world described by Lenaerts, in the early days the CJEU initially seemed to err on the side of ‘construction and safeguard’, drawing a clear line between the EU and national and international law to define the internal constitutional space of the Union. It was seen in the previous chapter that in Van Gend en Loos the CJEU famously ‘cut the cord’ with the legal orders of the Member States and effectively qualified the EU as a ‘tertium genus entity,’ so negating the ancient regime dichotomy between national and international law. It emerges from that decision that autonomy as ‘construction and safeguard’ arguably was originally an element to ensure the survival itself of the Union from external influences of national and international law: in the absence of a specific territory, polity, or government, on which the laws of each legal system apply, the way in which the ‘peoples’ are brought together was initially the subjection to the same ‘law of the land,’ which marked the confines of the EU legal system. That approach of ‘construction and safeguard’ thus initially filled the content of the constitutional order and its autonomy with some well-known tenets of EU law, including primacy, direct effect, general principles, and effective judicial protection, intended as ‘limitations’ to the operativity and scope of national and international law. Through the doctrine of primacy and direct effect, for instance, the Court construed the EU legal system as a hierarchical organisation of sources of powers, a ‘pyramid’ on top of which sit the EU treaties as a ‘tool of validation,’ defining the confines within which the EU can exist as a separate entity from both its Member States and international law. An early case on the operation of these principles can be found for instance in Dairy Products, where the CJEU rejected that the general principle of international law of the ‘exceptio non adempleti contractus’ codified in article 60 of the Vienna Convention on the Law of the Treaties (VCLT) could be relied upon in the EU to exclude that a failure on the part of the EU Institutions to comply with an obligation under the EU treaties – in this case, the obligation of the Council to set up an European market organization for dairy products to replace the national market organizations – could justify incompliance by Member States with their ‘standstill’ obligation under article 12 of the treaty of Rome. Judicial protection played a particularly important role in the early definition of the autonomy of EU legal system as ‘separation’ by adding some ‘bite’ to the principles of primacy and direct effectiveness of its laws: for the EU to act as an autonomous constitutional legal order the common law needs not only to exist, but also to be enforceable. Probably well conscious of the experience of its contemporary peers in international bodies and organisation, whose operation was heavily dependent on the willingness of the participating states, the CJEU seemed since the very beginning particularly preoccupied to ensure that EU law was enforceable as a matter of law, rather than as a gracious concession of the Member States. After all, the Court had made it abundantly clear in Van Gend en Loos that the EU was a different creature of international law. To that aim, the CJEU initially resorted to an ‘alliance’ with the national courts of the Member States, regarded as the ‘natural forum’ to safeguard the EU law-derived rights of their citizens. Interestingly, the CJEU based that alliance on a legal duty of national courts to exercise “direct and immediate protection” for their citizens found in their own common constitutional traditions and in articles 6 and 13 of the ECHR, which required them to offer a remedy against all breaches of rights. That duty was thus initially framed somewhat independently from EU law: the CJEU only invited national courts to collaborate – on their own terms – with the CJEU to safeguard an additional set of rights that Member States had conferred on their citizens by choosing to participate in the EU treaties. Judicial protection so responded to minimum standards of protection of EU law which enabled EU law to be enforced, and to operate autonomously from national and international law. So, in Johnston, a case relating to free movement of workers, the Court held that the “existence of a remedy of a judicial nature against any decision of a national authority refusing the benefit of that right is essential in order to secure for the individual effective protection of his right.” Similarly, in a series of preliminary opinions delivered under article 218(11) TFEU, analysed later in this work the Court, rejected the compatibility with EU law of numerous dispute settlement mechanisms envisaged in international treaties of the EU, on the ground that certain their characteristics would have impaired the functioning of certain basic mechanisms of the EU, particularly the ones giving effect to the principle of judicial protection. Arguably, these decisions of the Court in relation to both national and international law can be linked to a need to affirm minimum boundaries for the existence of the EU as an autonomous legal entity of a constitutional nature. Yet, based on markedly internal consideration of primacy of EU law and loyal cooperation, the general principle of judicial protection was progressively transformed by the Court in a much more invasive tool through which the Court managed its relationship with national courts and national legal orders. Firstly, in a key to curb differences among different national procedures, judicial protection in the EU had to respond to standards of ‘equivalence’ and ‘effectiveness’ compared to the protection offered at the national level. According to the Court, national rules governing judicial proceedings adjudicating on EU rights “must be no less favourable than those governing similar domestic actions” and should not render the exercise of those rights “practically impossible or excessively difficult.” It followed from this that, for instance, judicial protection must ensure that individuals must have actual access to independent and impartial courts to protect their EU law-derived rights. Similarly, the duty of sincere cooperation, found today in article 4(3) TEU, also required that national courts interpret and apply procedural rules governing national judicial proceedings in a way that achieves effective protection of EU-derived rights. The operation of judicial protection was further aided by the Mangold line of case law that general principles may, recurring certain circumstances, produce direct effect. Secondly, from Unibet – and even more evidently with Impact – the CJEU started enacting a more invasive scrutiny on enforcement of EU law by the national courts, requiring that its national counterparts shall ‘actively’ cooperate to secure primacy and unity of EU law by means of ‘effective’ remedies that actually deliver on the fundamental right of access to court and uphold the rule of law on the territories of the Union. In this framework, it was no longer sufficient for the Court that national law did not make it “virtually impossible or excessively difficult” to exercise EU law-derived rights. In the absence of tools that deliver on the effective enactment of EU law, new powers and remedies were to be created. Finally, with the Lisbon treaty effective judicial protection evolves into a tool through which the Court to bring the European integration process to a new level. Enshrined in articles 19(1) and 47 CFREU (Right to an effective remedy and to a fair trial), judicial protection is transformed from a general principle presiding to an obligation of results (ensuring the effectiveness of EU law) into an invasive primary law right underpinning an obligation of means (ensuring the effectiveness of EU law by complying with certain, EU-specific, standards of the rule of law) of national courts. That emerges in that, for the Court, it is no longer sufficient that all EU-law derived rights need to be effectively protected at the national level. In order to deliver on the essential premise of the EU legal system of the principle of mutual trust between judicial bodies of the Member States they also need to be subject to ‘same degree’ of protection across the legal system of the Member States. Significant cases that well describe this shift include Associação Sindical dos Juízes Portugueses, where the Court reviewed certain austerity measures reducing the remuneration of Portuguese judges from a standpoint of preservation of their judicial independence; a line of cases on the European Arrest Warrant relating to the surrender of individuals on human rights grounds; as well as Indépendance de la Cour Suprême, where the Court reviewed the reform of the Polish Supreme Court that subjected the age of retirement of the members of that court to a discretionary decision of the executive from the standpoint on the principle of irremovability from office. Significantly, in all these cases the Court makes of certain ‘substantive qualities’ of the national courts themselves borrowed by the common framework of Union’s values in article 2 TEU, rather than of their ‘procedures,’ the essential element to preserve the unity of the Union. Interestingly, that protection is also made more effective by emphasising its nature as an objective ‘characteristic of the EU legal order’ – enshrined in article 19(1) TEU – rather than a subjective element of protection of ‘individual rights’ in article 47 TFEU – whose application is more complex by virtue of article 51 CFREU and needs to be counterbalanced against any other competing rights. In this new regime, the approach of ‘minimum invasiveness’ and ‘deference’ towards national procedural autonomy of the origin seems to be effectively replaced by an approach of full hierarchical superiority of the EU primary law right to effective judicial protection. It thus unsurprising that the Court has in time shown to be particularly apprehensive of both national and international interferences with the functioning of judicial protection. Yet, the markedly ‘primacy’ rationale of the ‘enhanced’ primary law right to effective judicial protection that emerges from the developments described above would lead to think that its application should have a rather more modest significance in the international relations of the EU, compared to the internal relation. Yet, it will be seen in the remainder of this Ph.D., that differentiation does not necessarily emerge from the case law of the Court.
3.2 The original approach of the EU treaties and the status of international dispute settlement mechanisms
This approach of ‘construction and safeguard’ was arguably initially supported by the EU treaties. While internal and external autonomy are both rooted in the same claim of the constitutional nature of the EU in Van Gend en Loos, in the early times the Court was mostly concerned with defining the internal constitutional interactions of the Union with the legal orders of the Member States. It was indeed through those interactions that the identity of the EU was originally shaped. As Eeckhout notes, the constitutional importance of the relationship between EU law and international law originally emerges as a subordinate element, functional to the development of the internal constitutional interaction with the Member States and has little autonomous significance: without the definition of the ‘internal’ confines of the EU, issues of interaction with international law remained mostly a matter for the legal orders of the Member States. Significantly, the qualification “of international law” soon ceased to complement the references of the Court to “the new legal order” announced in Van Gend en Loos. The treaty of Rome contained very few provisions on the relationship between the EU and international law: the only powers conferred upon the EU in that sphere pertained to the common commercial policy – which functioned as a natural complement to the creation of the common market – and certain provisions on association agreements. The original provisions were limited to broadly set the perimeter of the field of competence of the EU in the external relations, but abstained from specifying the purpose for that action. That is, to an extent, still the case today: despite the broadening of the external objectives of the EU, which will be discussed later – unlike for the internal relations, the external action of the EU is generally defined in its arrival points rather than in the means to get to them. Arguably, the original approach of ‘construction and safeguard’ of the legal order from international law in the EU treaties and the Court’s case law also mirror the status of dispute settlement mechanisms at the time and the exceptionalism of the EU and its system of judicial protection. When the EU was created, there were few international courts, which functioned very differently from the Union’s model of compulsory jurisdiction, individuals’ protection, and real enforcement mechanisms. These courts of ‘first generation’ were the outcome of the process of started with the Jay treaty and the International Convention for the Pacific Settlement of International Disputes of 1899 (Hague Convention I), of creating institutions apt to uphold the rule of law through the facilitatation of the peaceful settlement of international differences when state diplomacy was unsuccessful. The first Hague Conference thus resulted in the creation of an arbitral institution available at all time to the contracting parties (the Permanent Court of Arbitration, “PCA”). The idea underlying the first (and second) Hague Conferences was that systems of rules could replace war as the main instrument of regulation of international relations. Yet, these courts and tribunals of ‘first generation’ for the most part had no compulsory jurisdiction, responded to the pressures of the litigating parties, were not open to private litigants, and did not possess leverage to induce governments to follow rulings they oppose. It is significant that the PCA is neither a ‘permanent’ nor a ‘court’ but rather only establishes a permanent secretariat and standing list of arbitrators to which the contracting parties to the Hague Conferences can resort in case disagreements arise. The individual-centred legal system of the EU and its mechanisms of judicial protection of the EU courts thus constituted a notable exception in the panorama of dispute settlement in international law. The approach of ‘construction and safeguard’ of the EU legal system from these mechanisms of outer scrutiny could thus also be read as an expression of the qualitative difference of the nature and functioning of the Union, as well as of the importance of full protection of the EU citizens.
4. a shift in international law: the new generation dispute settlement mechanisms
The context in which the EU and its unique dispute settlement mechanism were created soon evolved. Today a large network of effective institutions of ‘second generation’ have been ‘delegated,’ much like the CJEU, vast authority to help enforcing international law and reviewing its application by states and international organisations. ‘Second generation’ courts and tribunals find a common trait of being expression of states’ commitment to effectively enforce the international rule of law: while their structure and functioning may vary greatly, these dispute settlement mechanisms are mostly provided with compulsory jurisdiction, are more independent from external pressures, produce binding outcomes that can be effectively enforced internationally, actively develop rules of international law, and constitute essential mechanisms of judicial protection for non-state actors.. Like the CJEU this ‘second generation’ courts and tribunals are thus much more than ‘facilitators’ of the peaceful settlement of disputes among states and turn into ‘global actors’ of international law. That is particularly the case, it is argued, for the case study for this Ph.D., investor-state dispute settlement mechanisms (ISDS): while the instrument was created only a few years after the CJEU, appearing for the first time in the 1959 Germany-Pakistan BIT, like the CJEU its characteristics of independence make it a precursor of the ‘second generation’ dispute settlement mechanism. Specifically, much like the CJEU, ISDS tribunals possess compulsory jurisdiction to set up disputes before them; present numerous of guarantees of independence and impartiality from the parties; allow direct access to individuals to a supranational dispute settlement mechanism to uphold substantive standards of protection of the rule of law; present profiles of protection of fundamental rights; and are surrounded with efficient mechanisms of enforcement of their decisions. These ‘second generation’ courts and tribunals not only surround the CJEU with a large number of peers set up to achieve similar objectives of protection of the rule of law across borders. They arguably also are exceptional tools to export internal values and principles and influence the development of international law. To begin with, second generation courts and tribunals are a phenomenon that cannot be overlooked for quantitative reasons: these courts and tribunals have multiplicated and today adopt a very large number of decisions compared to the early times of the ‘first generation’ dispute settlement mechanisms. Alter notes that before the end of the Cold War there were only seven international courts, which, without counting the decisions of the CJEU, had collectively issued only 373 binding decisions. Today, the author counts at least twenty-four permanent international courts that issued over 37,000 judgments. Similarly, during its first 70 years of life the caseload of the PCA was very modest, comprising 25 arbitrations and 3 conciliations. These numbers are in stark contrast with the no less than 150 arbitrations administered by the secretariat between 2000-2014. That is particularly true for ISDS. The table of cases published by UNCTAD evidences that the number of investment arbitrations globally has grown exponentially in the past 30 years. The success of these dispute settlement mechanisms compared to the ‘first generation’ courts and tribunals can arguably be connected to their feature of giving standing to individuals. It is significant that, as Alter notes, in order to make the Andean legal system more effective, the contracting government authorized private actors to complain to the Secretariat, and to bring noncompliance suits to the (at the time) Andean Tribunal of Justice directly. Statistics published by the PCA and the Secretariat of the International Centre for Settlement of Investment Disputes (ICSID) in the past 20 years also show that disputes between individuals and states, intergovernmental organization, or other public entities consistently are by far the most numerous cases administered by these institutions. Interestingly, some of these courts and tribunals offer larger standing to individuals to start disputes compared to the one afforded to them in the EU. That is once again the case of the case study of this Ph.D. of ISDS. ISDS – traditionally included in many international investment agreements, contracts, or national law laws – are characterised by the fact that they allow a foreign investor to bring a dispute against a state with which its home state has an investment agreement in place. Secondly, second generation international courts and tribunals possess an array of functions which necessarily intertwines with international affairs of states and international organisations in the wider world (for instance, by influencing the interpretation of international rules) and internally (for instance, in case of enforcement actions on their territories). Notably, much like the CJEU, these courts and tribunals possess, in addition to their dispute settlement role, a variety of other roles, which Alter categorises in (i) enforcement of compliance with international law; (ii) administrative review of the decisions of administrative actors in cases raised by private litigants; (iii) constitutional review of assessing the legal validity of legislative and government actions vis-à-vis higher order legal obligations. Again, ISDS are a very good example of the different functions of the second generation courts and tribunals, due to their very effective tools of enforcement; their capacity to review under a broad number of substantive standards of protection host states’ decisions in relation to investors; and even the stability and fairness of the legal framework of host states. For this reason, investment arbitration has been regarded as “a form of global governance,” or as a “species of Global Administrative Law.” In addition to this, as Alter argues, independent international dispute resolution mechanisms often act as “tipping points” in the international relations, in so that they are global actors designed to influence, through their activity of adjudication, state preferences in the direction inscribed in their DNA. Economic courts will thus tend to pursue market openness, while human rights adjudicators will emphasise the protection of fundamental rights, and so on. With the result that “to the extent that international agreements codify the goals and objectives associated with economic liberalism or liberal democracy, [international courts] will more likely than not be a force contributing to these goals.” Therefore, participation in the definition of the DNA of new international courts can arguably impact on the direction that these courts will follow in adjudicating on international matters. Thirdly, these courts and tribunals today widely promote an individual-based idea of the rule of law shared by numerous states and international organisations. That emerges primarily in their main distinguishing factor compared to traditional dispute settlement of allowing access to dispute settlement to non-state actors – private litigants and/or supranational prosecutorial bodies – to initiate litigation (e.g., ISDS, Andean Tribunal). Secondly, second generation courts and tribunals tend to increasingly include the protection of individual rights in their decision-making. Once again, ISDS mechanisms are exemplary of this aspect. Notably, like the CJEU, investment arbitration is the outcome of the post-war efforts to “create equal rules for all.” ISDS arbitration indeed grants private foreign investors an unseen direct access to international arbitral tribunals, only comparable to some human rights courts, allowing them challenge alleged state restrictions on their rights under investment treaties and claim damages for their breach. Thanks to a large process of treaty reform, which will be discussed in detail later in this work, ISDS tribunals are also increasingly taking into account fundamental rights considerations in their decision making activity. The important role of ‘second generation’ courts and tribunals is further enhanced by their particularly effective tools to stimulate compliance with international law. Their higher independence in terms of conferral of powers, procedure, and ex post outcome of their decisions supplements the traditional instrument of ‘shaming’ states into compliance with the imposition of sanctions and effective procedures able to effectively deter states from deviating from their commitments. For instance, their compulsory jurisdiction reduces states’ power to decide when to initiate litigation as well as the risk of retaliations, incentivising them to fulfil their mandate of upholding the rule of law even against states’ will. Allee notes that referral of disputes to the 1965 Convention for the Settlement of Investment Disputes between States and Nationals of Other States (ICSID Convention) represents the highest degree of ‘delegation’ by states to an international dispute settlement mechanism, but the same seems true for most ISDS mechanisms: deferral to investment tribunals activates a procedural mechanism capable of arriving to decisions on the basis of pre-determined rules on which states have had little or no influence. For instance, the ‘ex ante’ state consent that characterises ISDS allows investors to start proceedings before arbitration tribunals immediately at each alleged breach. Most sets of arbitration rules also include various powers that allow tribunals to carry out the procedure independently by the parties, including visits of places and inquiries to non-parties or “default” procedures in the case that one of the parties attempts to stall the proceeding by not participating in it. This limits the traditionally greater power of states over individuals to derail or stop the arbitration proceeding. In addition to this, most sets of arbitration rules allow only very limited mechanisms of ex-post control over investment tribunals’ decisions. Recognition and enforcement of investment awards are ensured almost globally by the ICSID Convention, the 1958 Convention on the Recognition and Enforcement of Foreign Arbitral Awards (New York Convention), and the national arbitration laws based on the 1985 UNCITRAL Model Law, as amended in 2006. The system set up in the ICSID Convention is particularly representative of this aspect of investment arbitration: article 52 states that parties are bound by the award, which the state parties agree to consider as a final judgement of the relevant state’s supreme courts, and that the award can only be subject to limited grounds of review. Notably, unlike the New York Convention, these grounds of review don’t even contemplate the possibility of incompliance on the basis of “public policy” grounds. All these enforcement mechanisms have traditionally contributed to creating largely non-intrusive oversight by the national courts, which has been regarded as almost automatic.
It is also notable that, in order to preserve their relationship with individuals and overall legitimacy, second generation courts and tribunals tend to actively develop the efficiency and coherence of international law through both system development and dialogue. Second generation courts and tribunals thus often operate as a network of promotion of international law, participation in which is key for states and international organisations to orient their interpretation. By means of their interpretative action, second generation courts and tribunals firstly are able to ensure the day-to-day application of the law, and actively develop legislation by filling gaps and adding to the parties’ understanding of what the law is. In so doing, they supply an element of growth to sub-systems of international law. For instance, Alvarez notes the emergence of an increasingly more coherent regime of investment protection not only is the outcome of more specific provisions of domestic and international law, but also emerges from the “ever more elaborate interpretations of relevant law rendered by international arbitrators sitting in investor-state disputes.” The described advantages of the engagement states and international organisations with this ‘second generation’ dispute settlement mechanisms discussed until here come however at a price. Delegation of powers necessarily entails some limitation of the sovereign powers, just like Member States had to limit their own sovereignty to the CJEU. On the one hand, the creation of international courts and tribunals is an expression of ‘consent’ to be bound by international law for the pursuit their interests. On the other hand, it constitutes a delegation of powers that subjects domestic legal systems to the outer scrutiny of independent external actors with the power to decide what international commitments truly mean, at the exclusion of national institutions. By applying independently (their interpretation of) the rule of law to specific disputes, international courts and tribunals also have the power to inject foreign values and considerations in domestic legal orders, overcoming national political barriers and creating legal change across borders. The example of how the interpretative activity of the CJEU guided the development of the constitutional system of the EU, building an EU-specific body of values and principles is exemplary of this point. International courts and tribunals thus arguably have the potential to alter not only interactions between the EU and other international actors, but also its own internal structure. Secondly, outer scrutiny is liable to both implement and limit key fundamental principles of domestic legal systems, such the one of judicial protection in the EU. Section […] above already discussed the key role that outer scrutiny and judicial protection performed in the EU since the inception. Access to remedy went on to become, in the form of the principle of effective judicial protection, a key fundamental principle of the EU legal system and a fully-fledged fundamental right in both the Charter of Fundamental Right of the EU and in the European Convention on Human Rights. While outer scrutiny of international courts and tribunals is liable to safeguard and export access to remedy for EU and individuals, it however can equally impact on its implementation within the EU legal system, by interfering with the important function performed by the internal mechanisms set up for that purpose. It is telling in terms of the perception of the Court of the constitutional importance of judicial protection for the functioning and development of the EU legal system that the safeguard of the judicial system of the EU is the most frequently cited ground to declare the incompatibility of EU law and international law when the Court makes use of the doctrine of autonomy. In those decisions, while recognizing the importance of international dispute settlement mechanisms in the framework of the conclusion of international agreements, the CJEU carefully stresses the importance that those dispute settlement mechanism don’t interfere with the EU system of judicial protection. According to the Court, “the guardians of that legal order and the judicial system of the European Union are the Court of Justice and the courts and tribunals of the Member States” and that the judicial system of the EU set out in the EU treaties is essential “to ensure that the specific characteristics and the autonomy of that legal order are preserved” through guaranteeing “consistency and uniformity in the interpretation of EU law.”
5. the policy approach of the EU to outer scrutiny
5.1 A shift of approach: arbitration and ISDS mechanisms
In balancing the competing considerations described by Lenaerts between ‘safeguard’ of their sovereign powers and engagement with dispute settlement mechanisms, the EU has increasingly leaned towards enhanced openness and integration with outer scrutiny its external policy. Notably, while having originally showed a preference for diplomatic dispute settlement mechanisms, from the 1990s the EU started to adopt a more open approach to international law outer scrutiny as a tool for the realisation of the international interests of the EU. Notably, the 2001 Laeken Declaration provided that: What is Europe’s role in this changed world? Does Europe not, now that it is finally unified, have a leading role to play in a new world order, that of a power able both to play a stabilising role worldwide and to point the way ahead for many countries and peoples? Europe as the continent of humane values, the Magna Carta, the Bill of Rights, the French Revolution and the fall of the Berlin Wall; the continent of liberty, solidarity and above all diversity, meaning respect for others’ languages, cultures and traditions. The European Union’s one boundary is democracy and human rights. The Union is open only to countries which uphold basic values such as free elections, respect for minorities and respect for the rule of law. Similarly, the European Security Strategy committed the EU to creating stronger ties with international law based on a more solid “international society, well-functioning international institutions and a rule-based international order” as well as to developing international law within the framework of the UN Charter. Today, the block is the creation and reform of international courts globally, evidencing how the Union views dispute settlement mechanisms as an integral part of its commitment to multilateralism in the international relations. The described approach of the EU is firstly evident from the very active participation of the EU in the General Agreement on Tariff and Trade 1947, first, and the World Trade Organization, after, which came with a much more effective dispute settlement mechanism provided with compulsory jurisdiction. In time, the EU has been one of the most active users of that mechanism. The EU has further actively led the talks for reform of the WTO, notably during the Doha Round, which led to the 2017 ‘Trade Facilitation Agreement’ to expedite transfer, release, and clearance of goods for developing or least-developed countries, and, more recently, to overcome the paralysis of the WTO Appellate Body. In this last instance, the EU has notably led the development of the new ‘Multiparty Interim Appeal Arbitration Arrangement,’ which will temporarily enable its parties to continue benefitting from the binding resolution of trade disputes through an appeal review of panel reports. It is interesting for the topic of this Ph.D. that the Union’s preferred instrument of dispute settlement in international law is arbitration. From the 2000s, with the Lisbon strategy, state-to-state arbitration clauses have been systematically included in bilateral trade and cooperation agreements concluded by the EU and have increasingly become more complex to mirror the increasing complexity of the Union’s trade strategy. Arbitration clauses were also accepted by the EU multilateral contexts such as the in relation to the 1994 Energy Charter Treaty and the 1998 UN Convention on the Law of the Sea (UNCLOS) through which the EU accepts compulsory and binding arbitration (although not the compulsory jurisdiction of International Tribunal on the Law of the Sea (ITLOS). The enforcement of international trade and investment rules through arbitration was also identified as a top priority in the second phase of the Lisbon strategy and the Commission’s 2015 “Trade for All” strategy, which paved the way to the conclusion of a “new generation” comprehensive free trade agreements (FTAs) between the EU and third countries coupled with powerful arbitral dispute settlement mechanisms. An example is the “quasi-judicial” mechanism for the settlement of disputes in the EU-Ukraine Deep and Comprehensive Free Trade Area (DCFTA), which notably includes a preliminary reference mechanism in relation to matters of regulatory approximation. The EU also initially showed an approach of openness towards ISDS. To begin with, while the Treaty of Lisbon conferred only recently to the EU the competence to conclude investment agreements, it is interesting that the EU did not opt out from the ISDS mechanism included in the Energy Charter Treaty. Further, despite the Union not being able to access the ICSID Convention, only open for signature to States, the EU has made its views heard on cases involving Member States by means of amicus curiae briefs. This initial approach of openness towards ISDS was in line with the support that the mechanism traditionally found with the Member States of the EU. Shortly after the Treaty of Lisbon conferred upon it exclusive competence in the field of foreign direct investments, the EU however started to lead a global process of reform of investor-state arbitration, from arbitral to fully-fledged judicial bodies. This process is widely regarded as an effect of a general ‘backlash’ against ISDS arbitration, coupled with a growing adversity of the Commission against that mechanism for its alleged incompatibility with certain aspects of EU law. Yet, while the EU may be overcoming ISDS arbitration, its central position in the multilateral effort to reform the instrument evidences how the Union today views effective dispute settlement mechanisms as an integral part of its commitment to multilateralism in the international relations.
5.2 ISDS Reform
ISDS reform is taking place in the EU at three levels, internally to the EU, bilaterally, and internationally. Internally to the EU, the Commission has been leading for some time a campaign to terminate investment agreements existing between EU Member States (‘intra-EU’ BITs) and review the compatibility of investment agreements concluded by Member States with third countries. As for the first aspect, already in 2006, the Commission’s Economic and Financial Committee affirmed that part of the content of intra EU-BITs had been ‘superseded’ upon accession and recommended a review for the need for these agreements by the following year. As only two Member States – Ireland and Italy – terminated all their intra-EU BITs, the Commission started infringement proceedings against five Member States in 2015. Following the Court’s finding of the incompatibility of intra-EU with the autonomy of EU law in Achmea, examined below, the Commission coordinated three political declarations dated 15 and 16 January 2019, in which the Member States of the EU declared their intention to “terminate all bilateral investment treaties concluded between them by means of a plurilateral treaty or, where that is mutually recognised as more expedient, bilaterally.” The termination was eventually signed on 5 May 2020 by 23 Member States. The Commission has also immediately after started infringement proceedings against two states, Finland and the United Kingdom (despite the latter no longer being a Member State), that have failed to participate in it. While the termination agreement does not encompass ISDS proceeding started under the auspices of the Energy Charter Treaty, that treaty is also currently being renegotiated by the EU at the multilateral level (see below). As for the second aspect, Regulation (EU) No. 1219/2012 of December 2012 (Grandfathering Regulation) sets up a procedure to assess the compatibility of both, existing and of newly negotiated, Member States’ BITs with EU law. While providing that, in principle, existing BITs of EU Member States will be “progressively replaced by agreements of the Union relating to the same subject matter,” the Grandfathering Regulation however still allows Member States to negotiate new agreements with third countries provided where the EU itself has no investment agreement with that third country and that they are authorised by the Commission to do so. At the bilateral level, the EU Commission is negotiating the inclusion of the new investment agreements of the EU of an ‘Investment Court System’ (ICS), as a first step in the project for global reform of ISDS mechanisms. The ICS is an innovative two-tier mechanism for the adjudication of disputes between investors and States formed of a ‘tribunal’ and an ‘appellate body.’ Such mechanism is ‘hybrid’ in so that it presents characteristics of both arbitral tribunals – in that, for instance, its panels lack permanence – and courts of law – in that, for instance, the ICS includes an appellate body, higher guarantees of independence and the impartiality of its members, and enhanced transparency of its procedures – without squarely falling in either category. The ICS mechanism has to date been included in some of the most important recently concluded investment agreements of the EU, including the Comprehensive Economic and Trade Agreement between Canada and the EU (CETA), and the Investment Protection Agreement (IPA) with Singapore. At the multilateral level, the Commission has ‘revived’ the idea of a judicialized multilateral investment court – debated at lengths in the 1960s – and is currently negotiating the creation of a fully-fledged international investment court, the s.c. ‘Multilateral Investment Court’ (MIC). That negotiation is currently taking place in the framework of the United Nations Commission on International Trade Law (UNCITRAL) Working Group III, ahtough all the agreements in the context of which the EU has negotiated an ICS already include a commitment by both parties to transition, in the longer term, to the MIC. The Commission is also currently seeking to renegotiate the Energy Charter Treaty and its ISDS mechanism to bring it in line with in the ongoing multilateral reform process in UNCITRAL.
6. the renewed treaty approach to outer scrutiny
The shift towards greater ‘interaction’ with dispute settlement mechanisms of the EU executive develops in parallel with a renewed treaty approach to international law and international dispute settlement.
6.1 International law and outer scrutiny in the EU treaties
Successive treaty amendments progressively support a move towards an understanding of international law as an element for the full realisation of the identity of the EU. In this framework, the participation of the EU in international dispute settlement mechanisms plays a special role. The specific characteristics of ‘second generation’ international courts and tribunals discussed above make them an essential backdrop to many of the constitutional values, external objectives, and principles of the EU. To begin with, independent and well-functioning dispute settlement mechanisms undoubtedly support the fulfilment of the treaty values of the rule of law and justice enshrined in article 2 TEU and fall within the pool of instruments that ensure “peace, security […] free and fair trade, protection of human rights.” That suggests a favour in primary law for participation in – and compliance with the decisions of – international dispute resolution mechanism to uphold the rule of law. Along this line, Lenaerts for instance notes that inter-judicial dialogue of fundamental rights, particularly with the ECtHR, is today required by the CFREU. Outer control is also key to the legitimacy of the actions of the EU Institutions. As Eeckhout observes, genuine enforcement of values and external objectives of the EU in the external relations “ought to be located at the international judicial level, because of the stronger guarantees of independence offered by an “external” judiciary.” Outer scrutiny is indeed arguably key to democracy in so that it allows decentralised control on the extent to which the EU Institutions act in accordance with the characteristics of the identity of the EU, and thus respect the conditions for their mandate and legitimise their actions. For this reason, Alter notes that independent courts and tribunals support the development of a society based on the rule of law. In so far as they render accountable public officials, Alter argues, international courts enhance the quality of political decision-making and promote democratic accountability. So that “even if judges are political actors, not truly neutral or even unbiased, they can still be seen as better decision-makers than politicians.” Allee also observes that judicialization is a strong political tool for consensus and legitimacy on the part of democratic states, whose constituents support a strong commitment to the rule of law. Notably, it was seen above that judicialization was the response of the EU Commission to the “backlash” against ISDS arbitration, which led to the process of reform that the EU is today championing internally and externally. The EU treaty objective of the prevention of conflicts and the strengthening of international security “in accordance with the purposes and principles of the United Nations Charter” notably set the EU on a path to ensure the peaceful settlement of international disputes in accordance with the mechanisms included in article 33 of the UN Charter. Independent and well-functioning dispute settlement mechanisms can also contribute to “the strict observance […] of international law” to which the EU is committed; constitute “multilateral solutions to common problems” under article 21(1) TEU; and represent a tangible expression of the “high degree of cooperation in all fields of international relations” of the EU for the pursuit of the list objectives included in article 21(2) TEU. Particularly significant among them are the preservation of peace and the creation of “an international system based on stronger multilateral cooperation and good global governance.” Zang, for instance, points out that international courts in relation to which states do not have complete control over the conferral of powers, procedure, and ex post outcome of international proceedings present numerous benefit for good global governance and compliance with the rule of law. The author notes that independent courts enhance state compliance with international law, limit pressures from their side over the decision-making through the imposition of sanctions, and provide procedures able to effectively deter states from deviating from their international commitments. Thus, independent dispute settlement mechanisms are normally well placed to signal a strong commitment of their contracting parties to the full execution of their obligations. As a confirmation of this point, a study on international arbitration from Allee finds evidence that more powerful countries express a clear preference for the construction of independent dispute settlement mechanisms as they also expect their firms to be the ‘aggrieved’ party in future disputes. From another standpoint, dispute settlement mechanisms can contribute to the “the development of international law” as provided in article 3(5) TEU. Hudson argues that dispute settlement mechanisms with compulsory jurisdiction facilitate the development of the legal regimes in which they operate. The author notes that compulsory jurisdiction ensures that dispute settlement mechanisms may be in charge of the day-to-day application of the law through their interpretative action, and so actively develop legislation by adding to the parties’ understanding of what the law is. In so doing, independent dispute settlement mechanisms constitute an element of growth for sub-systems of international law. That was for instance evidenced above in the central role of the CJEU, as a catalyst of opposing interests, in the development of the EU legal system. Independent dispute settlement mechanisms also are regarded by Alter as ‘actors of change’: seeking to root their legitimacy more in the relationship with the beneficiaries of their decisions than in the approval of the states that set them up, they are incentivised to unhinge traditional rules and relations of power to promote compliance with the rule of law internationally. The discussed direct relationship that the CJEU sought with the individuals since Van Gend en Loos is a good example of this phenomenon. To a lesser extent, the Dispute Settlement Body (DSB) set out in the Dispute Settlement in the Dispute Settlement of Understanding (DSU) of the WTO also has the effect to give voice to less powerful countries, eliminating the ‘positive consensus’ element for the adoption of the reports of the arbitration panels. Further, access to independent courts and compliance with their rulings is also normally regarded as an essential element of the rule of law, which article 21(1) TEU lists among the elements that have inspired the Union’s “creation, development and enlargement, and which it seeks to advance in the wider world.” Goldstein and others argue that the creation of independent and well-functioning dispute settlement mechanisms constitutes a ‘long term commitment’ to the rule of law on the part of states: through them, governments bind their successors to compliance with a certain set of international rules for the future. The idea is that, while international law accepts withdrawal of consent from participation in these mechanism from the government of the day, that withdrawal will be subjected to a procedure and will be more likely to attract democratic scrutiny on its motivations. An example of this phenomenon is the membership of the United Kingdom of the Council of Europe. Despite its, at times, uneasy relationship with the Strasbourg court, evidenced for instance in the debate around the decisions of the ECtHR on prisoners’ rights, the governments of the day face great political internal and international pressures to remain part of that international organization. Arguably, participation in the creation of international courts is a key tool for the EU to also safeguard the values, fundamental interests, security, independence, and integrity of the Union outside of its borders, and of the reach of the jurisdiction of the CJEU, to advance its values and interests in the wider world. As discussed above, on the one hand, participation by the CJEU in the judicial dialogue allows the European Court to influence the development of international rules to which the EU is subject. On the other hand, participation of the Union in the definition of the DNA of new international courts can arguably impact on the direction that these courts will follow in adjudicating on international matters.
6.2 The traits of a new identity
EU treaties to an essential element of the constitutional identity of the EU is the introduction in primary law – through the treaty of Lisbon – of ‘values’ that expressly are said to apply to all Union action and of a broader set of ‘external objectives’ for the achievement of certain value-based purposes. These innovations and their significance form the basis for a new phase of European integration, and will now be examined in turn. While much of the content of EU values appeared already in previous codifications of the EU treaties, it is in the treaty of Lisbon they are for the first time qualified as such. This change constitutionalises a profound transformation in the identity of the EU that Sarmiento well depicts as a move from a “community of interest” into a “community of values.” The codification of EU-specific values in article 2 TEU is particularly significant in so that values are normally regarded as a typical element of statehood. That is even more notable considered the position they are given in the EU treaties – where they are included before the ‘objectives’ of the EU, traditionally regarded as a typical element of international organisations. For the purpose of this section, the fact that the provisions in articles 3(5) and 21(5) TEU provide that EU values constitute an element of the external action of the EU is significant not only because it provides for an overall coordination of internal and external EU policies. It is also significant because it acknowledges the existence of an external dimension of the common identity of the “peoples of Europe” behind which those peoples can unite. Turning to external objectives of the EU, it is significant that, as Lenaerts and Van Nuffel note, the new list of external objectives after Lisbon focusses “on non-economic goals to a far greater extent than the EC Treaty.” These objectives, which are spread across various provisions of the EU treaties, include contributing to “peace, security, the sustainable development of the Earth, solidarity and mutual respect among peoples, free and fair trade, eradication of poverty and the protection of human rights, in particular the rights of the child,” as well as to the consolidation and support of “democracy, the rule of law, human rights and the principles of international law,” and the promotion of “an international system based on stronger multilateral cooperation and good global governance.” The significant overlap and complementarity among certain of these external objectives with the internal objectives of the EU is evident. Suffice it here to take the example of ‘sustainable development,’ which is both an internal and external objective, or refer to the complementary nature of the creation of an area of “freedom, security and justice without internal frontiers” and the external objective of preservation of peace, prevention of conflicts and strengthening of international security. Taken together with the constitutional nature of the EU and the common set of values that underpins the internal and external relations of the EU, the new external objectives of the EU enshrine into primary law a relationship of interdependence between the internal and external dimension of the EU. One in which collaboration with third countries and participation in mechanisms set up to uphold the rule of law internationally is essential for the purpose of the fulfilment of its ideological blueprint. EU values and external objectives arguably fill the conditions and content of the constitutional relations of the EU with international dispute settlement mechanisms of new significance: not only they enshrine in the fundamental law of the EU that the block must maximize dialogue and engagement with them in order to implement its ideological blueprint. Significantly, they also define the manner of that engagement, which is a principled approach of cooperation and openness. In so doing, as Larik notes, values and external objectives ‘thicken’ the constitutional order of the EU adding qualitative traits and step up its international ambitions. Proactive and open engagement with the external world constitutes the ‘natural progression’ of the European project towards a new phase based on solid values, particularly the rule of law, and an element of what the block intends to be, not only to do. As internal autonomy, external autonomy thus arguably becomes in the EU treaties another tool for unification of the “peoples of Europe”: this time not through integration, as in the internal relations, but through identification behind a common European ‘way of life.’
6.3 The relationship with outer scrutiny of international dispute settlement mechanisms as an element of the identity of the EU
The arguments developed so far arguably support the argument that values and external objectives make of openness towards outer scrutiny of international dispute settlement mechanisms an element of the identity itself of the EU, which must guide the actions of the EU Institutions. Admittedly, the significance and legal nature of values and external objectives of the EU is highly debated. Their ‘constitutionalization’ in primary law has indeed been considered by some as “meaningless,” because of the intrinsic vagueness and of the level of abstraction they entail. That values and principles have a broad scope is indeed unquestionable. EU values, as enshrined in article 2 TEU, have accurately been defined by Von Bogdandy as “normative convictions of a highly abstract order […] part of the social identity of the individual” with a normative founding function for the whole of the Union’s legal order. The values of democracy, rule of law, and fundamental rights have also been defined by Timmer as “famously elusive” concepts. Like values, the external objectives of the EU, primarily enshrined in articles 3 and 21 TEU also have a very broad scope. As Cremona notes, unlike certain ‘internal objectives,’ such as the creation of the internal market, ‘external objectives’ tend to be much more ‘abstract,’ in so that they do not prescribe a specific action to be undertaken by the EU Institutions and the Member States but set up open-ended policy aims. The progressive broadening of the scope of external objectives to encompass non-purely economic aims, such as the objective to contribute “to peace, security, the sustainable development of the Earth,” arguably makes them more undefined than previous ‘economic’ external objectives, such as the “contribution to the development of international trade.” Yet, regardless of their broad scope, values and principles arguably have become today an element of the identity itself of the EU, which inevitably shapes the activity of its Institutions. The starting point of the reasoning is the ideological blueprint of the EU set out in the EU treaties of the creation of the “ever closer union.” That blueprint confers to the external objectives and values in the EU qualitative characteristics that make them different from the way other entities pursue their interests in international law and provides them with specific importance, as an essential part of its identity. As Manners maintains in his work on “normative power Europe,” not only are objectives and values primary norms of EU law, but they also have a deep constitutional significance as a step further towards the “ever closer union” by filling that program with significance and specific requirements. As they thus are part of what the EU is – and not merely of what it aims to be – all EU Institutions are bound to take them into account in their activity. Manner also notes that, as constitutional provisions underpinning the external action of the EU, those values and external objectives “predispose [the EU] to act in a normative way in world politics.” Put differently, values and external objectives set the broad lines of the EU international actorness. So, on whichever competence one chooses to set at the basis of its action, arguably the EU as a whole cannot act differently from what its constitutional requirements provide and should only act in accordance with them. Indeed, while they certainly don’t command a specific course of action, values and external objectives have been on many occasions interpreted and applied to specific cases, by the CJEU itself, and by other European Institutions to frame and justify the exercise of their powers. Values are for example considered ‘qualitatively different’ norms that shape the identity of the EU in the case law of the Court on external autonomy, which refers to them as “the very foundations of the Community,” or the “fundamental premiss” of the EU legal order. Furthermore, values and external objectives can be afforded a normative dimension. For instance, in so far as the value of human dignity in article 2 TEU finds concrete expression in article 1 CFREU, analogously titled, it ceases to be an abstract or ethical consideration, and becomes a fully-fledged right. Compliance with fundamental rights is indeed a legal requirement, subject to the scrutiny of the Court, and a condition of the lawfulness of EU acts. The value of the ‘rule of law’ also becomes a directly enforceable “politico-legal benchmark” in two provisions of the EU treaties, article 7 and 49 TEU: the procedure under article 7 TEU allows to impose sanctions on Member States that are found in “serious and persistent” breach of the rule of law. Article 49 TEU sets out a procedure for the accession of third countries to the Union, which can be initiated by “[a]ny European State which respects the values referred to in Article 2 and is committed to promoting them” (emphasis added). In that key, article 49 requires that the EU Institutions assess the compliance with EU acquis and the Copenhagen criteria of candidate countries and EU countries against which a rule of law procedure has been activated. The essential function of the elements of the identity of the EU as ‘fundamental structures’ of its action predicated, among others, by Manner, has important consequences for their role in the constitutional system of the EU. As primary law norms, values and external objectives cannot be regarded as mere ‘soft law’ instruments. On the opposite, as Eeckhout notes, their role in the constitutional identity of the EU, makes values and external objectives “meaningful to judicial discourse”, and, to an extent, enforceable. The author notes that the progressive constitutionalisation of EU law makes values and external objectives progressively relevant in a variety of fields of EU law. Indeed, it was seen above that it is the uncontested case law of the CJEU that both values and objectives, internal as much as external, are interpretative instruments that should orient the Union’s action. In particular, the abidingness of treaty objectives was affirmed on numerous occasions by the CJEU which also provided that that they “must be read together if they are to be properly applied.” The Court also specifically provided that the external objectives of the EU are an element that the Institutions must take into account in the implementation of their policies. As Larik notes, external objectives have been considered in the case law of the Court as a tool to both legitimise and inhibit external action. The common values in article 2 TEU have also on various occasions been relied upon by the Court as an interpretative tool to ascertain the compatibility of international dispute settlement mechanisms with the “very foundations of the European Union legal order.” Their intrinsic dynamic nature, which evolves together with the constitutional structure of the EU, makes of values and external objectives features in constant evolution. Notably, in Opinion 2/13 the Court observed with regards to the Union’s objectives in article 3 TEU that “those provisions, which are part of the framework of a system that is specific to the EU, are structured in such a way as to contribute — each within its specific field and with its own particular characteristics — to the implementation of the process of integration that is the raison d’être of the EU itself.” Therefore, according to the Court, the conditions and content of the Union’s interactions with international law must reflect the elements of the identity of the EU as it evolves in time. The nature of values and external objectives as elements of the identity of the EU indeed makes them dependent on the dynamic nature of the EU legal system on the basis of the dialogical evolution of the EU discussed in chapter […]. As a result, as the constitutional EU legal order ‘thickens’, values and external objectives add to the overall considerations that the EU Institutions, Court included, need to take into account and balance in their action, making them more complex. That was, for instance, the case of fundamental rights, which, as from Internationale Handelsgesellschaft, need to be “protected within the framework of Community law.” From that standpoint, values and external objectives can be regarded as legal rules in progress which orient law-making and policymaking. Arguably, values and external objectives also have the additional function of obliging the Institutions to continuously pursue them in the exercise of their powers, orienting them towards a certain direction.
In a way, the progressive emergence of the identity of the EU constitutes the natural outcome of a path towards the fulfilment of the ideological blueprint of the EU of the creation of the “ever closer union.” A path made of progressively more specific commitments that the EU undertakes, as part of its identity, to make its own in all areas of action and which are used to progress and to justify its action. As Larik notes, the connection between values and objectives and justification for the power and role of the EU is most evident in the draft articles for the Draft Constitution, where the Præsidium explain that its aim was “to set out the general objectives justifying the very existence of the Union and its action for its citizens.” The CJEU also appears to confirm this reading where it connects the values in article 2 TEU to the foundations and purpose of the EU. It follows from the above that, while their broad content admittedly cannot create outright obligations on the EU Institutions to act in a specific manner, values and external objectives can certainly limit their discretion by requiring them to justify both the manner in which they act externally and the way they receive external influences internally. In turn, incompliance with external objectives and values may give raise to issues of legitimacy of the exercise of powers.
7. setting the parameters for a renewed approach of the court to international law and outer scrutiny
The strong commitment of the EU executive towards outer scrutiny supported by the EU in the external relations, together with progressive endorsement of the EU treaties of a more open and dialogic relationship with international dispute settlement mechanisms as an element of the constitutional identity itself of the EU, should arguably lead to reassess the Court’s original understanding of external autonomy. Two main considerations contribute to this conclusion. On the one hand, the fact that, as a ‘doctrine’ of constitutional interaction, autonomy can only develop within the framework of the EU treaties. Therefore, the approach of ‘safeguard’ of the origins needs to evolve to take reflect the very different considerations that today preside the international relations of the EU. A ‘stronger’ and more confident Union, surrounded by powerful and effective international courts, no longer needs adopt an approach of separation from international law to affirm its autonomy. Rather, it advances its international agenda through active engagement with others on the basis of its values and principles. On the other hand, the more ‘passive’ role of the CJEU in the definition of the external autonomy, of ‘protector’ and ‘defender’ of the EU from international law. As Cremona notes, the CJEU has traditionally not been an outright positive “driving force behind the EU’s policy agenda” as in the internal relations, where it traditionally was a ‘catalyst’ and ‘mover’ of the EU legal system. That is still true today. As Cremona observes, the EU treaties indirectly set limits to the role of the CJEU in the external relations by conferring in that sphere a more prominent role to the political Institutions of the EU. The reason being that the external relations of the EU, developed among equally ordinated sovereign entities, traditionally require some scope for discretion to be able to balance the different interests underlying external action and find compromise in the achievement of potentially competing external objectives. That space for discretion for the political Institutions of the EU, it is noted here, especially includes the policy decisions regarding the conditions and content underpinning the relationship between the EU and other international actors – i.e., the scope of external autonomy. The highly political nature of the international relations requires indeed a careful balancing of the efficient conduct of the external relations with the fundamental values of the EU in article 2 TEU and its external objectives, that only the political Institutions and not the EU courts, can carry out. For that reason, the EU treaties define a specific role for each EU Institution. It is noteworthy that the parameter of the control of the CJEU arguably depends on the nature, more or less political, of the power exercised. For instance, the Court has a more limited scrutiny in relation to CFSP matter compared to the scrutiny that it is allowed to it under the CCP. As Lazowski and Wessel note, the rationale for this arrangement was arguably “to prevent judicial activism in this area of EU competence.” That approach reflects a fundamental issue of institutional balance common to many systems of law, on the basis of which foreign relations are traditionally subject to special rules and their regulation is mostly reserved for the executive. The idea is that, in order to advance their policy objectives in the international relations, Governments require some space for discretion, diplomacy, and negotiation. For this reason, foreign relations are often qualified as a domain for ‘high politics.’ Reasoning on the idea that this discretion of Governments is necessary for them to carry out their tasks and preserve the separation of powers between the different branches of the state, courts have traditionally applied a lower scrutiny of judicial review – or even refrained from exercising any – to the domain of foreign relations. Institutional balance in the external relations plays thus a fundamental role in the external relations and is thus regarded as a constitutional principle of EU law. Hillion evidences that the principle of institutional balance in the EU pursues two main constitutional functions: on the one hand, it preserves the integrity of the institutional framework in the EU treaties. On the other, it ensures that the institutional framework renders the Union able to fulfil its tasks. Institutional balance is thus an essential element to live up to its ideological blueprint of the EU of creating the “ever closer union,” which arguably underpins considerations of internal as much as external autonomy. From another standpoint, unlike internal action, the external action of the EU expands beyond the confines of the EU legal system, assuming a legal significance and producing legal effects that partly fall outside of the jurisdiction of the CJEU. For instance, the international treaties between the EU and third countries are both EU and international acts with potentially different application and interpretation under EU and international law. The EU treaties also provide for a markedly ‘inward-looking’ jurisdiction of the Court. Articles 256, 263, 267, and 270 TFEU, which include the main heads of jurisdiction of the Court, refer to actions between Institutions and Member States, both among themselves and against each other; exceptionally, individual actions against regulatory acts of the EU; preliminary references from national courts; and actions between the Union and its civil servants. Article 218(11) TFEU gives the CJEU jurisdiction over the ‘internal’ questions of compatibility of international treaties of the EU, prior to their conclusions, with its constitutional framework. Article 275 TFEU expressly limits the scope of the Court’s jurisdiction over the Common Foreign and Security Policy. For all these reasons, in the external relations the Court has hardly developed a constructive and positive connotation as a tool that affirms identity, as it has also done in the international relations. As it is evidenced in its limited jurisdiction in the external relations, the Court rather has a role of ensuring that international law may not impinge in the constitutional project of unification being carried out internally through the formation of an identity of the EU. Elevating openness and dialogue with international law and dispute settlement mechanisms to the position of primary law arguably even broadens the discretion of political and executive Institutions of the EU in carrying out the external policy and accepting compromise with international law. Considering the highly political nature of the external relations and the specific institutional balance in that sphere, one would thus expect the Court to carefully exercise the large discretion intrinsic to the concept of autonomy in a way which respects as much as possible the “space for voluntary acceptance of external norms” of the other Institutions. That includes their voluntary acceptance of the cession of some of their sovereign powers to set up international dispute settlement mechanisms that contributes to fulfilling the ideological blueprint of the EU treaties. Just as some sovereignty had equally to be given up by the state parties to the ECSC in exchange for the advantages arising out of that treaty. Taken together with the treaty approach of openness to outer scrutiny this specifically means resolving the tension underlying the doctrine of autonomy between separation to build the constitutional system and integration with international law in the sense of principled support for the Union’s participation in international dispute resolution mechanisms, only ‘rebuttable,’ on the basis of a case-by-case analysis.
8. conclusions
Chapter […] analysed the development of the interactions of the EU with international law and outer scrutiny, with an eye specifically on international arbitration and ISDS mechanisms. Section […] identified the existence of two sides of autonomy, internal and external. The section observed that, while the former concerns the constitutional interactions between the EU and its Member States, the latter regards the constitutional interactions between the EU and international law. Section […] also clarified that this Ph.D. will focus on external autonomy, taking as a case study the constitutional interactions between the EU and dispute settlement mechanisms carrying out outer scrutiny over it – particularly arbitral tribunals. The chapter then moved to analyse the development of the policy and constitutional framework underpinning the relationship between the autonomy and outer scrutiny. Section […] firstly observed that the original approach of the Court of ‘separation’ of the EU from outer scrutiny, often applied in a key to safeguard the judicial system of the EU from fragmenting interferences of national and international law with fundamental elements of the EU constitutional legal order, particularly judicial protection. That approach, it was argued was supported by the inward-looking nature of the original EU treaties and reflected the perception of the EU and its Courts as an exceptional, superior, tool in international law to uphold the rule of law compared to other courts and tribunals adjudicating on disputes across borders. Yet, that situation soon changed. Section […] observed that the emergence of a ‘second generation’ of international dispute settlement mechanisms has today been delegated the authority to effectively support international actors to uphold the rule of law across borders. That ‘second generation’ courts and tribunals are provided with guarantees of independence and effective tools of enforcement of their decisions that are similar, and sometimes go beyond, the ones conferred to the CJEU. They thus have transformed from ‘facilitators’ of the settlement of disputes into powerful instruments of global governance. That is particularly the case of ISDS tribunals, which arguably possess compulsory jurisdiction, allow large standing to individuals, uphold rule of law standards that are universally recognized, are able to carry out administrative, and even constitutional review of the legal framework of respondent host states, and possess powerful tools of enforcement of their decisions. While engagement with this second generation courts and tribunals is thus arguably key to participate in the global discourse shaping international law, it comes at the price of giving up some sovereignty. Section […] observed that, after some initial perplexities, the external policy of the EU increasingly reflected acceptance of such limitations and today envisages international dispute settlement as an integral part of the EU commitment to multilateralism. Evidence of this is the position of the EU as a leader and a mover of a vast process of creation (in FTAs) and reform (notably, WTO, ISDS, and ECT) of such mechanisms. A shift towards a position of favour towards outer scrutiny is similarly reflected in the EU treaties. Section […] observed that a favour in primary EU law for participation in – and dialogue with – international courts and tribunals, not only is today the backdrop of numerous primary law provisions of the EU treaties. It also has become a fundamental element of the identity itself of the EU, essential to live up to its ideological blueprint of the realisation of the “ever closer Union” for the benefit of the peoples of Europe, and should thus guide the actions of all EU Institutions, CJEU included, in the external relations. Section […] evidenced that one would thus expect to see the same considerations prevail in the case law of the EU on autonomy. On the one hand, autonomy needs to develop within the framework of the EU treaties. On the other hand, while the Court has always possessed a role of ‘protection’ of the EU legal system in the external relations, the larger policy space conferred to the political Institutions of the EU in that sphere invites an approach of restrain on the part of the Court in the exercise of the discretion intrinsic to the doctrine of autonomy. That restrain, it was argued, should concretise in particular attention to safeguard the space of voluntary cession on their part of sovereign powers for the creation of dispute settlement mechanisms that contribute to fulfilling the ideological blueprint of the EU treaties. For these reasons, section […] concluded, in shaping the doctrine of external autonomy the CJEU should adopted a principled approach of openness towards outer scrutiny and embrace the rule of a low degree of scrutiny of the compatibility of dispute settlement mechanisms with EU law. The next chapter will now move to analyse the approach of the Court to outer scrutiny in several cases in which the Court has decided on the compatibility with EU law of international dispute settlement mechanisms.
CHAPTER IV – THE COURT’S INTERPRETATION OF THE AUTONOMY AND OUTER SCRUTINY
1. Introduction
The previous chapter put forward the argument that the EU executive and the EU treaties today privilege active engagement of the EU with international courts and tribunals over an approach of protection of the EU legal system. In particular, that engagement acquires in the EU treaties not only the value of an essential tool for the Union to achieve its aims in the wider world, but, arguably, of an element of its identity itself. The CJEU should thus in principle adopt a lower degree of scrutiny regarding the compatibility with EU law of dispute settlement mechanisms. This chapter looks at how the CJEU has in practice developed the doctrine of autonomy in relation to international courts and tribunals, taking as a case study the approach of the Court to arbitral tribunals, and particularly ISDS arbitration. The chapter is organised as follows. Section […] delineates the scope of the analysis carried out in this work and sets out the reasons for the choice of the case study. Section […] analyses the preliminary opinions of the CJEU under article 218(11) TFEU on the compatibility of EU law with international courts and tribunals, with particular focus on Opinion 2/13, regarding the compatibility with EU law of the Union’s accession to the ECHR. Section […] looks at various aspects of the complex relationship between the CJEU and arbitral tribunals, with special focus on international investment tribunals. The conclusions of this chapter are discussed in in the next chapter […].
2. Scope of the analysis: outer scrutiny of courts and arbitral tribunals
While the Court’s understanding of external autonomy discussed in chapter […] is very broad, this Ph.D. focuses on the narrower topic of its interface with courts and tribunals carrying out outer scrutiny over the EU, taking arbitral tribunals, particularly ISDS tribunals, as a case study. The rationale for the choice of this topic reside in the vast array of issues that it underlies. As discussed, the characteristic of ‘second generation’ international courts and tribunals to act as actor of global governance sits uncomfortably with a view of autonomy of EU law as fully preserving self-determination and independence, as it necessarily entails giving up some of its sovereign powers. In particular, outer scrutiny of international courts and tribunals naturally limits the full operativity of the system of judicial protection of the EU, in so that either it provides alternative routes of protection outside of that system or interferes with its functioning. That is particularly true for arbitration. Arbitration is indeed traditionally viewed as a more flexible, swifter, dispute resolution mechanism compared to national courts, which allows the parties to decide their disputes by means of tribunals whose composition, language, and applicable law can be agreed by the parties. Its widespread use as an alternative to the use of state courts also multiplies the opportunities for conflict with the system of judicial protection of the EU. The hybrid nature of arbitral tribunals as dispute settlement mechanisms only partially integrated with national legal systems makes their compatibility with the autonomy of EU law and its system of judicial protection of the EU particularly complex. On the one hand, a private mechanism of dispute settlement, arbitration is often not fully separate from national courts and relies on them in relation to ‘ancillary matters’ essential to its functioning, such as the enforcement of awards or the appointment of arbitrators. Yet, arbitration may function separately from state courts as an autonomous system of dispute settlement, and the rationale for its selection by the litigating parties is often found precisely in that separation. That hybrid nature interestingly can make of arbitral tribunals, at times, key tools for the functioning of the system of judicial protection of the EU (when fully integrated with it), and, other times, potentially dangerous interferers with the judicial system of the EU (when it functions separately from it). For these reasons, while admitting that the interest to efficient arbitration proceedings in principle limits the possibility for national courts to review arbitral awards, the CJEU has oftentimes intervened to adjudicate on the extent to which arbitration may bring litigation outside of the judicial system of the EU at the expenses, in particular, of the monopoly of the CJEU on interpretation and application of EU law (344 TFEU) and the preliminary reference mechanisms (267 TFEU). ISDS arbitration has to date given rise to the some of the most relevant issues of compatibility between the autonomous nature of the EU and arbitration. The approach of the Court to the specific category of investment arbitration is particularly relevant for this Ph.D. for two reasons. Firstly, temporal. It is significant that the competence on foreign direct investment, was only introduced in the treaty of Lisbon, which also incorporated the definitive shift of the EU approach towards openness and dialogue with international law in the identity of the 27-block. The second phase of the Lisbon strategy, developed during the same years, also signalled a strengthening of the EU policy approach of openness towards international law through the conclusion of broader international agreements coupled with powerful dispute settlement mechanisms. One would thus expect the case law of the Court on investment arbitration to reflect this shift. Secondly, investment arbitration as a particularly powerful form of ‘global governance’ raises all sorts of interesting questions on the right balance between the ‘safeguard’ of the autonomy of the constitutional structure of the EU, particularly of its system of judicial protection, and openness towards outer scrutiny of a dispute settlement mechanisms that can so effectively protect of individuals’ rights – set at the centre of the EU legal system. While some of these questions were already assessed by the CJEU, arguably many others are yet to be analysed. The first problematic issues already in relation to ISDS and autonomy of EU law is that direct access of investors to ISDS tribunals without prior exhaustion of domestic remedies creates a route to bring litigation outside of the national courts of the Member States, impacting on the functioning of the EU system of judicial protection in the form of the preliminary reference procedure under article 267 TFEU, and the monopoly of the interpretation and application of EU law on the part of the Court. The issue was analysed by the CJEU in connection to the intra-EU’ BITs discussed above. The second issue arising from ISDS arbitration is that this mechanism a) upholds different standards of the rule of law compared to EU ones; and b) allows individuals to directly challenge state measures against those different standards, potentially interfering with the autonomous standards of protection set out in EU law. Like the CJEU, investment tribunals are set out to uphold the rule of law in international disputes. Yet, the differing standards of protection upheld by investment tribunals – especially when coupled with their broad jurisdiction also on in situations regulated by EU law – create a potential conflict with the EU-specific rule of law. The CJEU has already found a conflict between EU and international standards of protection upheld by investment tribunals in relation to the EU rules on free movement of capitals – in the form of articles 64, 66, and 75 TFEU – with the ‘transfer clauses’ included in BITs signed by Member States with non-European countries. On numerous occasions, the Court observed that, by guaranteeing investors and their subsidiaries established in the EU the unconditional right to transfer capital related to investment, these clauses unduly restricted the power of the EU to regulate how capital is moved between third countries and Member States. While not been expressly discussed by the CJEU, one such particularly problematic overlap of standards of protection regards the protection of fundamental rights, and particularly the right to access to court and full judicial protection. On the one hand, fundamental rights have historically constituted an important ground for the CJEU to consolidate and expand EU competence and its own jurisdiction, on the grounds of their ‘horizontal’ nature, which cuts across subjects. While, as Sarmiento recalls, not initially designed to be a human rights court, and still today rejecting that connotation, the subsidiary protection of human rights that the CJEU offers has played in the years an important role in broadening its jurisdiction and strengthening its relationship and legitimacy vis-à-vis individuals, making of EU courts a one-stop forum for the full judicial protection of their rights under EU law. Notably, today the EU has provided itself with its own standards of fundamental rights protection, codified in the CFREU, and, as it was evidenced by the Court’s approach to the Union’s accession to the ECHR (and to Opinion 1/17 below), it has been keen to ensure that other standards of protection external to the EU – be them national or international – don’t unduly interfere with its own. On the other hand, the broad territorial and subject-matter scope of the jurisdiction of ISDS tribunals – progressively regarded as an avenue also for the protection of fundamental rights – shapes and enforces a set of international human rights standards potentially competing with the EU-specific ones. Much has been written about the relationship between ISDS mechanisms and fundamental rights. Notably, international investment treaties and their precursors – the friendship, commerce, and navigation treaties – are considered the earliest conceptions of human rights in international law. Both ensured “legal protection of individual rights by means of legal and judicial restraints on government powers, such as prohibitions on discrimination, substantive safeguards of individual freedoms and private property rights, and judicial remedies at national and international levels, protecting individuals against abuses of power.” Interestingly, these agreements are creatures of the same movement towards the valorisation of individuals rights developed after the end of world war II that is at the origin of the EU. As St Jones interestingly evidences, before moving to promote the conclusion of an international treaty for the protection of foreign investments that would have become the ICSID Convention, the ‘father’ of ISDS arbitration, Aaron Brochers, was himself a victim of illegal acts of foreign countries. Having found refuge from the Netherlands in the United States at the outbreak of the II world conflict from the persecutions against Jews, Broches returned to its birthland to find that his family had been exterminated by the Nazi regime and that his properties, including his father’s cigarette business, had been illegally expropriated. Broches eventually filed a successful claim for compensation for the illegal expropriations. It is perhaps not a case that he then made the objective of his life to build a system featuring a right to access to justice before independent tribunals for investors finding themselves in the same situation. Today, the main standards of protection of investment agreements, expropriation, non-discrimination, and even a right to an effective remedy and to a fair trial (e.g., the Fair and Equitable Treatment (FET) clause – a standard of protection that requires states to make decisions in good faith – has been regarded as encompassing a variety of guarantees of the right to judicial protection, such as transparency, stability and act consistently of the legal framework, principle of legality, denial of justice, and ensure due process) find corresponding protection in the ECHR and the CFREU. Investment tribunals also have based their decisions on human rights standards included in international treaties, and notably entered in judicial dialogues with human rights courts on the definition of standards of protection. The characteristic of ISDS arbitration of allowing individual access to dispute settlement has also been regarded as problematic by the EU Commission in so that it potentially curtails the ‘right to regulate’ of the EU. According to the Commission, by allowing direct challenges on the part of investors to EU law or state measures implementing EU law for the enforcement of standards of protection competing with EU-specific ones, ISDS arbitration would interfere with the Union’s ability to regulate for the benefit of its citizens: the idea is that the obligation to pay compensations ordered by a tribunal, or even the threat of potential litigation. may bring the EU or its Member States to act or refrain from acting in the public interest, curtailing their political mandate. Indeed, ISDS provides investors with much broader standing rights than those conferred to individuals under EU law, allowing them access to arbitration, normally without the prior exhaustion of national remedies, on the basis of the ‘original’ consent given by states (or the EU, in relation to EU investment agreements) when including arbitration clauses in the investment treaties. That is the case, for example, of the Energy Charter Treaty, to which both the EU and its Member States are parties. Under that treaty, investors of a contracting party can ‘accept’ the offer to arbitrate of the state parties to the ECT by bringing disputes before arbitral tribunals constituted ad hoc based on the alleged breach of one of their rights under that treaty.
3. Analysis of the case law
3.1 Degree of scrutiny in preliminary opinions
The preliminary opinions under article 218(11) TFEU appear particularly relevant in a key of analysing the understanding of autonomy of the Court in relation to international courts and tribunals generally for two main reasons. Firstly, after Costa, the Court expressly discussed autonomy of EU law was the preliminary opinion in Inland Waterways. From that moment onwards, most of the cases in which the Court expressly discusses autonomy are preliminary opinions under article 218(11) TFEU. Secondly, the large timespan covered by these opinions gives an overview of the development of the approach of the Court to autonomy and outer scrutiny. Because this Ph.D. uses as a case study outer scrutiny of arbitration tribunals, with the only exception of Opinion 2/13, this section will only assess the general degree of scrutiny in preliminary opinions adopted by the Court in relation to autonomy and international courts and tribunals. Article 218(11) TFEU provides for the possibility for the Commission, the Council, the Member States, and, after Lisbon, the European Parliament, to request an opinion to the Court on the compatibility with EU law of international treaties being negotiated by the EU ahead of their conclusion. The idea is to avoid forestalling complications resulting from legal disputes concerning the substantive or procedural compatibility with the treaties of EU international agreements and prevent difficulties arising in the internal and international relations of the EU. The instrument of preliminary opinions under article 218(11) TFEU has been activated seldomly. The online database of the Court shows that, overall, this jurisdiction of the Court has been triggered in 21 cases since the conclusion of the treaty of Rome, with the first one being decided in 1975. The Court has understood the scope of article 218(11) TFEU broadly, accepting requests not only on the compatibility of ‘envisaged’ international treaties with EU law, as the wording of the article would appear to suggest, but also in relation to the more markedly ‘internal’ matter of the scope of the Union’s powers to conclude international treaties. These latter are the most frequent (13 out of 21), with only eight opinions focussing on the compatibility of features of envisaged international treaties. In so far as they allow the judiciary to take decisions before the arousal of an actual conflict with the EU legal system, the preliminary opinions of the CJEU broadly recall the tool of ‘advisory opinions,’ an instrument to which common law is traditionally adverse for reasons of separation of powers. The reasoning being that it is the place of parliaments, and not of judges, to pronounce on the meaning of the law in abstract terms. Courts should instead more properly apply the law in an adversarial context, where specific rules are to be applied to define rights, duties, or liabilities. These considerations could be regarded, to an extent, as equally applicable to opinions under article 218(11) TFEU: the anticipatory nature of that procedure, which provides that the Court decides on the text of an agreement before its implementation, means that the incompatibilities assessed by the Court are often only potential and future. They thus confer large discretion to the judiciary in an area, that of the foreign relations, which was seen to be naturally dominated by the executive. That is all the more true in so far that the preliminary opinions of the CJEU under article 218(11) TFEU are not ‘advisory’ but rather provided with legally binding force: where they are adverse, either the envisaged international agreement is amended, or the EU treaties revised. The scrutiny of the Court thus entails a particularly high risk of invading the fields of competence of the executive relating to the framing and implementation of the international policy of the EU and upsetting the institutional balance set out in the EU treaties. On this basis, one would expect the Court to adopt an especially low threshold of compatibility of international treaties with EU law in the delivery of its opinions under article 218(11) TFEU. On the surface, the CJEU has accepted the need for this lower degree of scrutiny. On numerous occasions the Court has expressly affirmed that the creation of mechanisms for the resolution of disputes in international agreements is part of the general competence of the EU to act in the external relations. Yet, the impression that one gains from reading the relevant opinions of the Court is that, while formally recognizing the importance of international dispute settlement mechanism for the correct functioning of the EU legal system, the Court in practice has consistently adopted a very high level of scrutiny of their compatibility with the constitutional structure of the EU and normally concludes against their legality under EU law. That is true not only for ‘older’ cases of the CJEU, decided at a time when both the EU treaties and the EU executive supported an approach of ‘protection’ of the EU from outer scrutiny. That approach rather remains constant thought the years and arguably reaches its peak when the approach of openness of the EU towards outer scrutiny was well-established in 2013, with the opinion on the EU accession to the ECHR. That conclusion firstly emerges from the numerical data that the Court has only upheld such compatibility only in three cases out of the nine in which it was called to decide on the issue: Opinion 1/92, relating to the creation of the EFTA Court, Opinion 1/00, relating to the system of legal supervision proposed by the Agreement, on the establishment of a European Common Aviation Area, and Opinion 1/17, on the investor-state dispute settlement mechanism set out in the Comprehensive Economic Trade Agreement (CETA) between the EU and Canada. Notably, while Opinion 1/92 already is a ‘second attempt’ to have the judicial system in the EEA declared compatible with the EU treaties, Opinion 1/00 on the Agreement on the European Common Aviation Area, only upholds the legality of a ‘Joint Committee’, thus a body which essentially solves disputes through diplomatic means, rather than a fully-fledged dispute resolution mechanism. Further, the Agreement on the European Common Aviation Area essentially ultimately extends the power of the EU Institutions to third countries by giving the contracting parties the option of allowing their courts or tribunals to refer questions to the Court for a preliminary ruling, and granting to the EU Commission the power to ensure that the competition rules of the European Common Aviation Area are complied with throughout that area. The high scrutiny of compatibility normally adopted by the Court in its opinions under article 218(11) TFEU is also confirmed by the fact that incompatibilities between EU law and international law are often phrased in terms of ‘possibility’: the Court appears to often satisfy itself of the existence of an incompatibility with EU law even in cases in which only a remote possibility exists that the dispute resolution mechanism set out in the scrutinised international treaty may sit uncomfortably with certain features of EU law. For instance, among other things, in Opinion 1/76 (Inland Waterways) the Court found that article 6 of the analysed draft regulation, providing that six members of the Court of Justice would sit on the Fund Tribunal, was incompatible with the EU treaties on basis as this created a chance that the selected CJEU judges “might be prejudicing their position as regards questions which might come before the Court of Justice of the Community after being brought before the Fund Tribunal and vice versa.” The Court continued in its reasoning suggesting that this situation may create a conflict with the duty of impartiality of its judges when matters related to the regulation come before the CJEU, as the Court “might” find “in extreme cases impossible to assemble a quorum of judges able to give a ruling on contentious questions which had already been before the Fund Tribunal.” Equally, in Opinion 1/91, on the EEA Court, the CJEU “feared” that the application of the provision seeking to create organic links between the EEA Court and the Court of Justice by providing organic links between the courts may make it “very difficult, if not impossible” for those judges, when sitting in the Court of Justice, to decide questions of law with completely open minds. The Court also deemed incompatible with EU law that the preliminary ruling it would have delivered to the EFTA states would have been purely advisory as this, again, “may give rise to uncertainty about their legal value for courts in Member States of the Community.” In Opinion 1/09, relating to the agreement for a Patent Court, the CJEU criticised the proposed international dispute settlement on the basis that that Court “may have been called upon to determine a dispute pending before it in the light of the fundamental rights and general principles of European Union law, or even to examine the validity of an [EU] act.” Opinion 2/13 on the accession to the ECHR, appears to include the most invasive degree of scrutiny of a dispute resolution mechanism included in an international agreement of the EU. Among other things, the Court decided that the mechanism established by protocol 16 of that agreement, which allowed the highest courts and tribunals of the Member States to request advisory opinions to the ECHR on issues relating to the interpretation or application of the rights and freedoms protected by the Charter, could affect the ‘autonomy’ of the preliminary ruling procedure in article 267 TFEU. The Court declared in that regard that “it cannot be ruled out that a request for an advisory opinion [...] might” circumvent the mechanism of preliminary reference. With regards to the mechanism of prior involvement of the CJEU in the decision-making activity of the ECtHR was also deemed incompatible with EU law on the basis that the agreement doesn’t “contain anything to suggest that [the] possibility is excluded” that the ECtHR may be allowed to give a judgement on issues on which the CJEU has already delivered a judgement. The Court eventually decided that these and other reasons stood in the way of the accession of the EU to the European Convention on Human Rights. The invasiveness of the scrutiny of the Court in Opinion 2/13 appears even more striking if one considers the political context in which the opinion was delivered: after its earlier Opinion 2/94,where the Court had rejected that the Community had the competence to access the European Convention on Human Rights, the treaty of Lisbon set out a precise obligation for the Union under article 6(2) TEU to do so. In addition to being at the odds with a treaty approach of clear favour for outer scrutiny of the EU, Opinion 2/13 effectively defies a precise EU treaty obligation. It does so on the grounds of conflicts that are not actual and impending, but merely potential. For all these reasons, Opinion 2/13 thus seem to show that the degree of scrutiny of the Court of the compatibility with EU law of international dispute settlement mechanisms even heightens in time. The high degree of scrutiny adopted by the Court in the context of the preliminary opinions under article 218(11) TFEU also emerges by the frequent express reliance of the Court on the doctrine of autonomy in these cases. Arguably, the characteristics of doctrines as theorization on the part of courts of what the law is to influence its future development, makes them best suited to cases in which the law does not offer to the Court a clear support for its decisions and the CJEU wants to resort to systemic assumptions to reach decisions which reframe or push the boundaries of the legal system through its precedent.
a. Opinion 2/13
Opinion 2/13 is an interesting example of the Court’s approach to outer scrutiny of international courts and tribunals: while adopted shortly after the treaty of Lisbon came into force, Opinion 2/13 adopts quite a strong stance on the degree of compromise that the EU can(not) accept with international dispute settlement mechanisms. The opinion even seems to signal a stiffening of the rigid position of the Court which sits uncomfortably with the development of an approach of openness to outer scrutiny set out in the EU treaties and endorsed by the EU executive. As a matter of background, the accession of the EU to the ECHR had been on the table since the 1970s. A first attempt in that sense was rejected by the CJEU in Opinion 2/94, adopted under article 218(11) TFEU on the grounds that the (at the time) Community did not have the competences to do so. The Commission started re-negotiating the accession of the EU to the ECHR in 2010, on the basis of a mandate conferred by the Council. Accession to the ECHR then became an obligation under article 6(2) TEU, which in the first part provides in express terms that “The Union shall accede to the European Convention for the Protection of Human Rights and Fundamental Freedoms,” included in the EU treaties as a response to the first (failed) attempt of the EU to access the ECHR. On 18 December 2014 the CJEU delivered its long-awaited Opinion 2/13 on the accession of the European Union to the European Convention on Human Rights. To the surprise of many, while the Court declared to be ready to accept that the EU may be subjected to outer control, it found that the ‘Draft revised agreement on the accession of the European Union to the Convention for the Protection of Human Rights and Fundamental Freedoms’ (Draft Agreement) was incompatible with EU law. The first argument of the Court was that the accession to the ECHR would breach the ‘specific characteristics and the autonomy of EU law.’ According to the Court, the accession to the ECHR would give “to the ECtHR the power to review the CJEU’s findings on the extent to which a Member State was bound by the EU’s own human rights guarantees under the Charter of Fundamental Rights of the European Union” or may rely on the Charter to escape its EU law obligations under article 53 CFREU. In so doing, the accession treaty would breach the autonomy of the EU to set up different, higher standards of protection of human rights on its territories. The Court was concerned that accession would breach the principle of mutual trust among Member States and “upset the underlying balance of the EU” by enabling domestic courts of the Member States to cast doubts on whether other Member States comply with their ECtHR obligations when implementing EU law. Further, the Court was unsatisfied with Protocol 16 to the ECHR which, by allowing the highest “courts and tribunals of the Member States to request the ECtHR to give advisory opinions on questions of principle relating to the interpretation or application of the rights and freedoms guaranteed by the ECHR or the protocols thereto,” set up a concurrent mechanism to the preliminary reference envisaged by article 267 TFEU and created a risk of circumvention of this latter procedure. The second argument of the Court revolved around the compatibility of the envisaged accession with article 344 TFEU. The Court reasoned that the fact that, upon accession, the ECHR would become part of EU law created the possibility that the EU or Member States might bring proceedings among each other for alleged violations of the ECHR in conjunction with EU law. That fact in itself went for the Court “against the very nature of EU law,” considering that such law “requires that relations between the Member States be governed by EU law to the exclusion, if EU law so requires, of any other law.” Thirdly, the CJEU was unsatisfied with the ‘co-respondent mechanism,’ which had been envisaged by the drafters of the accession treaty to “ensure that proceedings by non-Member States and individual applications are correctly addressed to Member States and/or the Union as appropriate”. According to the Court, articles 3(5) and 3(7) of the Draft Agreement would allow the possibility that the ECtHR could decide on the proper respondent in each case by assessing the plausibility of the reasons given by Member States and the EU when requesting to intervene in judicial proceedings before the ECtHR; by reason that Member States could be held responsible as co-respondents for violations of rights in relation to which they have expressed reservations in article 57 ECHR; and because of the possibility that the ECtHR may in any case find only one of them responsible for a violation. Further, the Court took issue with the procedure for its prior involvement in disputes before the ECtHR. The Court argued that this was due to the both the fact that nothing in the Draft Agreement excluded that the ECtHR could decide whether the Court had already adopted a decision “on the same question of law as that at issue in the proceedings before the ECtHR;” and that that mechanism would have not allowed the CJEU to decide on the interpretation of secondary law. Finally, the CJEU found that the arrangements contained in the Draft Agreement in relation to judicial review in CFSP matters did not preserve the “specific characteristics” of the EU legal system. The Court reasoned that, by empowering the ECtHR “to rule on the compatibility with the ECHR of certain acts, actions or omissions performed in the context of the CFSP, and notably of those whose legality the Court of Justice cannot, for want of jurisdiction, review in the light of fundamental rights” the Draft Agreement would have substantially entrusted “the judicial review of those acts, actions or omissions on the part of the EU exclusively to a non-EU body.” While most of the arguments of the Court admittedly had already been developed in previous cases, the overall sense of the Court in Opinion 2/13 is an enhanced high degree of scrutiny in the external relations of the EU from numerous standpoints, and has been for that reason subject to a great deal of criticism. That firstly emerges from the fact that some of the mechanisms found problematic by the Court – including the co-respondent mechanism and the prior involvement mechanism – had been specifically designed by the negotiating parties to address the issues arisen in its previous case law of the Court on autonomy. Yet, in Opinion 2/13 the Court does not consider them ‘good enough,’ both showing a very high degree of scrutiny in the conduct of the external relations by the Institutions and evidencing the lack of clarity of the guidelines set out in the jurisprudence of the Court on autonomy. Secondly, as previously mentioned in relation to the general analysis of preliminary opinions, in Opinion 2/13 the Court makes full use of its ‘language of possibility:’ the opinion is filled with caveats, ‘ifs and ‘buts’, which bring into question the possibility itself that the doctrine of autonomy of the CJEU may be compatible with any form of outer scrutiny on the EU.
Moving to the analysis of the case, it is firstly unclear from Opinion 2/15 how the ECtHR would bring into question the findings of the CJEU “in relation to the scope ratione materiae of EU law” in order to determine whether Member States are bound by fundamental rights of the EU. On the one hand, it is difficult to see on what grounds the ECtHR would have decided on the scope of application of the CFREU under its article 51(1). On the other hand, the ECtHR would also have not been able to evaluate how the level of protection of a fundamental right by a Member State compares to the level of protection of that same fundamental right in the EU, as interpreted by the CJEU. All the ECtHR could have done in case of accession is to decide whether one of those two standards complies with the degree of protection of rights under the Convention. That outer scrutiny is the rationale itself of the accession of the EU to the ECHR and the assumption that it would threaten the ‘autonomy’ of the EU would translate into an outright incompatibility between autonomy of the EU and any international dispute settlement mechanism. Equally unclear is how article 53 CFREU would have impinged in the autonomy of the EU by allowing Member States to adopt higher standards of protection of fundamental rights, so infringing the unity of EU law. As Eeckhout also notes, while the wording of article 53 ECHR does allow such a possibility, in so that it safeguards the right of the parties to the Convention to set up a higher degree of protection of fundamental rights, that possibility remains very remote. Its use by the Court as an argument to declare the incompatibility of the Draft Agreement with EU law arguably is evidence of a very high standard of scrutiny. The reason being that, once again, it is unclear how the ECtHR could, through its limited jurisdiction on the compatibility of the degree of protection of the Member States and the EU of a fundamental right vis-à-vis the Convention, require Member States to adopt a higher level of protection of fundamental rights on their own territories. The result of this interpretation of article 53 ECHR thus appears to exclude tout court the compatibility of the accession of the EU to the ECHR. The further critic of the CJEU that the Draft Agreement would not have safeguarded the principle of mutual trust by allowing Member States to bring actions among themselves and against the EU before the ECtHR equally points towards a high degree of scrutiny of its compatibility with EU law and sits uncomfortably with the approach of openness and dialogue toward international dispute settlement mechanisms set out in the EU treaties. Notably, by fencing off intra-EU actions from the scrutiny of the Strasbourg Court, the CJEU essentially demanded a reduction of the scrutiny that the ECtHR currently already carries out on the breach of fundamental rights among Member States. In so doing, the Court shows to be prioritising the principle of mutual trust over the effective protection of fundamental rights without there being any indication of that hierarchy of values in the EU treaties. Indeed, while mutual trust is admittedly one of the principles that form the very structure of the EU, the safeguard of fundamental rights is part of the substance of the identity itself of the EU. For that reason, they stand at the forefront of European integration. It is almost paradoxical that the protection of human rights, initially integrated in the EU system of law to lower resistance of national courts to the implementation of EU law, is now being used by the Court to fence off the scrutiny of an international body into EU law. That approach certainly proves a “certain puzzlement in the Union’s supreme jurisdiction approach towards human rights and their institutional implications.” The concern of the Court that the preliminary involvement procedure included in Protocol 16 to the ECHR would raise issues of autonomy of EU law also is quite difficult to grasp. To begin with, as Advocate General Kokott also pointed out in her opinion, Protocol 16 was not part of the Draft Agreement, but was negotiated and signed independently from it, and independently from that Draft Agreement it will come into force. So that it is doubtful that the Court even had jurisdiction to decide on its compatibility with EU law in the first place. It is also unclear how that procedure could ‘interfere’ with the prior involvement mechanism included in the Draft Agreement on the request of the CJEU itself. While the involvement of the CJEU in the context of a preliminary reference by the national courts and tribunals of the EU would certainly have been complex in terms of handling, it is unclear from the reasoning of the Court how it could have impinged in the fundamental characteristics of the EU. This further argument of the Court also seems based on a very broad understanding of irreconcilable conflict between EU law and international law that sits uncomfortably with the framework of the EU treaties. The reasoning of the Court on the issues raised by the Court in relation to article 344 TFEU also strikes for its high degree of compatibility with international law for two main reasons. On the one hand, it is yet another example of the CJEU considering a mechanism set out by the negotiating parties to safeguard the autonomy of the EU that is considered by the CJEU as not being ‘good enough:’ article 5 of the Draft Agreement indeed expressly safeguarded the exclusive jurisdiction of the Court by providing that that jurisdiction was not to be considered one of the mechanisms that the contracting parties to the ECHR have agreed to forego under article 55 of the Convention. On the other hand, the CJEU made quite an uncompromising statement by expressly excluding that its jurisdiction under article “344 TFEU precludes any prior or subsequent external control.” While the exclusive jurisdiction of the Court on EU law set out in that article cannot be questioned, it is unclear on what treaty basis the CJEU rules out that the EU Institutions may allow an external body to review the compatibility of the case law of the Court against an external comparator such as the ECHR. Considering that the ECtHR is not called to assess the decisions of the CJEU against the EU treaties, its outer scrutiny certainly would not compromise the right of the Court to make its decisions. For that reason, it would not impinge in the unity of EU law that article 344 TFEU protects. The scrutiny of the ECtHR is indeed limited to the powers of any mechanism of external control, i.e., an assessment of whether EU law as a whole is compatible with a certain source of international law. Deeming that control incompatible with the EU treaties equates to tout court excluding that the EU Institutions may enter in any negotiations setting up an international dispute settlement mechanism. The reasoning of the Court is difficult to follow in so that, as Eeckhout also notes, it finds against the possible jurisdiction of the ECtHR on a subject-matter – the compatibility of the treaties against a source of equal status – on which the Court itself does not itself have jurisdiction. While it is debatable whether it would be advantageous for the EU to expose itself to potential challenges of the Member States in relation to the compatibility of its primary law with the Convention, that choice surely must lie with the Council rather than with the CJEU. Finally, one last point, also raised by AG Kokott in her opinion, appears relevant for the purposes of this analysis: by adopting a broad interpretation of article 344 TFEU as ruling out all possible intra-EU state to state dispute settlement mechanisms, the Court, probably inadvertently, ends up ruling out the compatibility of longstanding international agreements of the EU providing for that arrangement, including, among others, those under the WTO agreements. While one would exclude that the Court was intending to generate such long-reaching and structural consequences, bringing the reasoning of the Court on article 344 TFEU to its extremes is helpful to evidence, once again, the extremely high degree of scrutiny that it exercises on the Draft Agreement in Opinion 2/13. From another standpoint, less than clear is also how the co-respondent mechanism, purposefully elaborated in the Draft Agreement to prevent the breach of the Court’s red lines on autonomy, would “not ensure that the specific characteristics of the EU and EU law are preserved.” Firstly, the argument is yet another confirmation of the high degree of scrutiny in the work of the ECtHR on the part of the CJEU. On the one hand, the Court was not at ease with the fact that the ECtHR may assess even the motivations put forward by the relevant Member State or the EU on the inclusion of co-respondents in the proceeding. On the other, it regarded as inadmissible that the ECtHR exceptionally may find at the end of the proceedings that only one of the co-respondents is internationally responsible for a breach of the Convention. In so doing the Court excluded that the ECtHR may have any say on the apportionment of the responsibility between the EU and Member States, which arguably is a crucial feature of any dispute settlement mechanism, essential to ensure the effectiveness of judicial protection of the claimants. For that reason, the approach of the Court once again equates to a tout court exclusion of any outer control on the EU legal system. Nor it is clear at all how the apportionment of international responsibility would give jurisdiction to the Strasbourg court on the internal division of powers among the EU and its Member States: arguably, the ECtHR would only be called to assess whether the act or omission allegedly violating the ECHR was mandated by EU law or was adopted by Member States in the exercise of their discretion in implementing EU law. The Strasbourg court would hardly need to decide on the division of competences in the founding treaties. Even if it that was the case, it is difficult to see how that review could ‘change’ or even meaningfully affect EU law, so to create an actual threat to its autonomy. Finally, as Lazowski and Wessel note, it is striking that the Court also rejected the compatibility of the Draft Agreement with EU law on the ground that that agreement would confer on the ECtHR jurisdiction to rule on certain issues of Common Foreign and Security Policy (CFSP), on which the CJEU has limited jurisdiction. The reasoning of the Court on this point evidences once again a too strict scrutiny on the choice of the EU Institutions to access the ECHR and almost gives the impression of a ‘selfish’ Court. The CJEU seems to effectively motivate its criticism on that, if itself cannot provide a remedy against the acts of the Institutions in the CFSP, then no other court shall be able to do so. Nor that argument may be justified on the grounds that the ECtHR could effectively breach the EU treaties by carrying out a judicial review of the acts of the EU Institutions against the EU treaties. It cannot be stressed enough that the only jurisdiction conferred upon the ECtHR relates to the compatibility of the acts of the EU Institutions with a source of international law which, by definition, sits outside of the EU legal system, i.e., the ECHR. In conclusion, the overall impression that one gains from reading Opinion 2/13 is one of increased unfriendliness of the CJEU towards international law, which manifests in a particularly high degree of scrutiny in the functioning mechanisms of the outer scrutiny set up in the Draft Agreement. The issues found by the CJEU with the mechanisms included in the accession treaty are so significant and require such radical review on the part of the negotiating parties to take account of peculiarities of EU law that it seems hardly possible that they could be implemented in a way that is acceptable to both of them. Even if they were able to find such compromise, the issue would remain of whether it is even possible to meet the demands of the CJEU and at the same time set up an international mechanism able to deliver meaningful outer scrutiny on EU law.
3.2 Autonomy issues with international arbitration
The sections below review some of the most significant decisions of the Court relating to the compatibility of various aspects of arbitration with EU law in a view of assessing the degree of scrutiny adopted by the Court in relation to various types of arbitration: commercial, state-to-state, and investor-state arbitration. Also in relation to these decisions, it would appear on the surface that the CJEU has accepted that the efficiency of arbitration proceedings calls for a limited review of arbitration awards on the part of national courts and should be possible only in exceptional circumstances. However, analogously to what it was seen above, in practice the Court often adopts a high degree of scrutiny of the compatibility of EU law with arbitration proceedings with the system of judicial protection of the EU, signalling an approach of ‘safeguard’ rather than openness towards these dispute settlement mechanisms.
a. Definition of “court or tribunal”
A first strand of cases exemplificative of the approach of the CJEU to the constitutional relationship between arbitration and EU law concerns the decisions of the Court regarding the possibility for arbitral tribunals to refer preliminary questions under article 267 TFEU. Allowing preliminary references from arbitral tribunals – whether deciding on the grounds of national or international law – would arguably implement the treaty commitment to openness and dialogue with international judiciaries and implement the international objectives and values of the EU. Allowing arbitral tribunals to make references to the CJEU would indeed reduce the potential for conflict between arbitration awards and EU law, enhance certainty in the application of the law across borders, and, ultimately, ensure more effective judicial protection for the users of both systems. From a purely symbolic standpoint, allowing preliminary references from arbitral tribunals would also be a clear sign of the willingness of the CJEU to dialogue with entities sitting outside of the EU. However, while the preliminary reference procedure has been successfully and effectively used to foster a dialogue between national judiciaries and the CJEU, the Court has traditionally declined the idea to entertain such dialogue with arbitral tribunals deemed to sit outside of the system of judicial protection of the EU. The rationale of this approach of the Court is debated, and it is certainly a consideration to keep in mind when analysing the case law that, as Sarmiento notes, any amendments of the preliminary reference mechanism are liable to change the face itself of the EU judiciary. Yet, whatever the rationale for that approach may be, the case law of the CJEU on preliminary reference from arbitral tribunals would seem to show that not only the Court has been quite reluctant to include arbitral tribunals within the definition of ‘court or tribunal’ under article 267 TFEU. By relying on a not always consistent interpretations of whether a tribunal may make preliminary references under article 267 TFEU, it would even seem to be increasingly reluctant to allow arbitral tribunals to refer preliminary questions.
The development of the case law
The preliminary reference mechanism was included in the ECSC treaty on the input of the Italian jurist Nicola Catalano, later a judge of the European Court of Justice, who had experience with the functioning of the Italian Court of Cassation. Under (what is today) article 267 TFEU, all courts and tribunals of last instance in the EU are under an obligation to ask the CJEU clarifications on any doubts that they may entertain on the interpretation of EU law. The underlying idea is that the CJEU should be entrusted with a tool for both securing uniformity of the law throughout the Member States and preventing the formation of diverging bodies of case law across their legal systems. It was seen above that the preliminary reference procedure became indeed a key tool for the construction and development of the EU legal system. In the absence of a clear legal definition in the EU treaties, the Court has developed its own criteria to ascertain what constitutes a “court or tribunal” under article 267 TFEU. In Vaassen-Göbbels, a case regarding a reference from an arbitration tribunal with an appeal function for decisions brought against the decisions of a social security institution, the CJEU laid out five determining criteria to decide whether a body constitutes a “court or tribunal” for the purposes of article 267 TFEU. These are the (i) statutory origin; (ii) permanence; (iii) inter partes procedure; (iv) compulsory jurisdiction; and (v) application of rules of law. In the later cases of Pretore di Salò and Corbiau the Court also added to the above the criterion of independence. It is evident how some of these criteria, especially those of permanence and compulsory jurisdiction, are liable to exclude from the outset arbitral tribunals. These criteria have been very flexibly applied in time. So flexibly that AG Colomer in De Coster, referred to the case law of the CJEU as being “casuistic, very elastic and not very scientific, with such vague outlines that a question referred for a preliminary ruling by Sancho Panza as governor of the island of Barataria would be accepted.” An analysis of the case law of the Court shows that the observation of AG Colomer is often not far from the truth. Not only the Court very flexibly decides on the application of some or all the mentioned criteria, but also it interprets them variably, in a manner that may at times be regarded as almost ‘result-oriented.’ The outcome is that the case law is often inconsistent, and it remains difficult to foresee which position the Court will adopt from time to time. Vaassen-Göbbels, mentioned above, is one of the first instances where the Court was called to decide whether an arbitral tribunal – the Scheidsgerecht van het Beambtenfonds voor het Mijnbedrijf – could be defined as a “court or tribunal” for the purposes of article 267 TFEU. The Court found that that was the case. The CJEU reasoned that the tribunal was constituted under national law (the Reglement van het Beambtenfonds voor het Mijnbedrijf), was permanent, decided in adversarial procedures, had compulsory jurisdiction (hear appeals brought against the decisions of a social security institution), and applied rules of law. Therefore, although the Dutch arbitration tribunal did not form part of the Netherlands legal system, both the Court concluded that arbitral body only had the appearance of a tribunal and it could be regarded as a court or tribunal for the purposes of the preliminary reference mechanism. The Court reached an opposite conclusion in relation to the arbitral tribunal Nordsee. The case regarded an arbitration clause included in a secret contract for the fraudulent division of EU funds among builders of factory-ships for fishing. The seat of the arbitration was Germany and the applicable law was German law. Pending arbitration, the arbitrator decided to stay proceedings and referred three questions to the CJEU. Before getting into the merits of the Court’s decision, it is relevant to note that, under German law, decisions of private arbitrators’ are substantially equated to courts’: arbitrators must decide on the basis of the law rather than equity and their awards are equivalent to final decisions of national courts. Yet, the CJEU declined jurisdiction stating that the referring tribunal could not be qualified as “court or tribunal” of a Member States under article 267 TFEU and that the similarities of arbitrators with judges in the case at issue were “not sufficient to give the arbitrator the status of a ‘court or tribunal of a Member State.” For the Court, the link between arbitrators and national courts was not sufficiently close as: (i) there was no obligation for the parties to resort to arbitration proceedings; (ii) Germany had not entrusted them the duty of ensuring compliance with its obligations under EU law, as evidenced by the fact that its public authorities were neither involved in the decision to arbitrate, nor allowed automatically intervene in arbitration proceedings. The CJEU thus concluded that questions of EU law could only be raised before national courts in proceedings parallel to arbitration proceedings. The decision in Nordsee seemed conclusive in the sense that if arbitration proceedings were compulsory and of last instance as a matter of national law the Court would consider the preliminary question. Indeed, in Danfoss the Court considered a preliminary question referred to it by a Danish arbitration court. The CJEU observed that the referring arbitration court was legally granted final jurisdiction in issues concerning collective agreements between organisations of employees and employers; that it had compulsory jurisdiction since parties could start proceedings despite the objections of the others; and that its judgements were final and binding. The Court seems here to adopt a more flexible reading of its own criteria of application of article 267 TFEU. Indeed, it overlooked that the jurisdiction of the Danish arbitration tribunal could be derogated by collective agreements – and thus was not compulsory. It also seemed to turn a blind eye to the fact that the composition and functioning of the Danish arbitration board was ad hoc and set out by the law only in its broad composition guidelines. The flexible understanding of “courts or tribunals” under article 267 TFEU is confirmed by many other cases. The inconsistency of the case law of the Court becomes particularly evident upon comparison of three significant recent cases on the definition of “court or tribunal”: Ascendi and Mereck Canada, on the one hand, and Achmea, on the other. Considering the similarity of the reasoning of the Court in the two cases, the discussion below will focus on the decision of the CJEU in Ascendi.
Ascendi
In Ascendi, the Court accepted a preliminary reference by a Portuguese arbitral tribunal, the Tribunal Arbitral Tributário set up to deal with certain taxation cases. The CJEU reasoned that it constituted an expression of the “procedural autonomy” of the Member State at issue to confer the jurisdiction for this type of dispute upon bodies other than ordinary courts. It appears not to be a coincidence that in both Ascendi and Merek Canada the referring tribunals were Portuguese. As the Court mentions itself in the decision, Portugal has a strongly monistic approach to international law and its legal system automatically integrates “norms and principles of general or common international law” including arbitration conventions such as ICSID Convention and the UNCITRAL arbitration rules. The country’s favour for arbitration is also expressed in other specific constitutional provisions, such as article 209(2), which states that “arbitral tribunals” are considered jurisdictional bodies. Building on its previous case law, the Court developed a full four-layered test to define what is to be understood as a court or tribunal under article 267 TFEU. However, a closer look to the application of such test appears to reveal a number of inconsistencies with the previous case law. To begin with, in relation to the criterion that the tribunal needs to be established by law, the Court observed that arbitral tribunals are included in the list of national courts under article 209 of the Portuguese constitution. For the Court, the legal rules governing tax arbitration qualify arbitration as an “alternative means of judicial resolution of tax” – which confer upon tribunals “general jurisdiction” to assess the “legality of the payment of any tax”. The reasoning of the Court on this point sits however uncomfortably with Nordsee, where the Court declined jurisdiction despite German law de facto assimilating arbitration to court proceeding. That is even more true if one considers that both in Nordsee and Ascendi the route of arbitration remained a mere option for the litigating parties, as the relevant arbitral tribunals in both cases lacked compulsive jurisdiction. Similar inconsistencies with the previous case law of the Court are found in relation to the criterion of permanence. The CJEU assumed the permanence of the Tribunal Arbitral Tributário from the fact that it was entrenched in the law. However, the CJEU failed to consider that the formation of the panels of the Tribunal Arbitral Tributário remains temporary as their activity terminates once an award is rendered. The Court deemed sufficient to tick the criteria of compulsory jurisdiction and procedure inter partes of the Tribunal Arbitral Tributário that its decisions bound the parties and that it had compulsory jurisdiction. The Court however overlooked the fact that ‘compulsory jurisdiction’ by nature does not belong to commercial arbitration, including the one before the Portuguese tribunal at hand. In a somehow circular reasoning, the Court observed that the jurisdiction of the Portuguese tribunal is not subject to the will of the parties – but only once the taxpayer applicant decides to submit the dispute to tax arbitration. The Court insisted on the fact that, from that moment on, the Tribunal Arbitral Tributário has, in accordance with the relevant national jurisdiction, compulsory jurisdiction as regards taxation and customs matters. As for the criteria of independence and application of rules of law, the Court arrived to the conclusion that the Tribunal complied with these characteristics as the appointment of the tribunal’s members was made by the Ethics Board of the Centre for Administrative Arbitration from among the list drawn up by that institution. The Court also looked at the fact that the existence of any personal or professional relationship between the arbitrator and one of the parties to the dispute was qualified by the law as an impediment to the exercise of the function of arbitrator. While one may agree with the arguments of the Court, it is relevant that the CJEU failed to acknowledge that most arbitration laws and agreements include comparable rules on independence of arbitrators, especially on conflict of interest.
Achmea
The cases in Merek Canada and Ascendi sit particularly uncomfortably with the decision of the Court of 6 March 2018 in Achmea. Reasoning on the “exceptional nature” of investment tribunals, in that case the Court decided that an investment tribunal such as the one established under article 8 of the Dutch-Czech Bilateral Investment Treaty (BIT) is not a “court or tribunal” under article 267 TFEU. As it is also apparent in the opinion rendered by Advocate General Wathelet in that case, the outcome of the reasoning of the Court that investment tribunal don’t possess the characteristics of courts or tribunals of Member States under article 267 TFEU was not so obvious. Interestingly, while the Court heavily relies on the Ascendi test to make its determination, as also observed by AG Wathelet in his opinion, a strict application of the Ascendi test to the investment tribunal in Achmea seems instead to support an opposite conclusion. The Court firstly distinguished Achmea from Ascendi on the grounds that the tax tribunal at issue in Ascendi was integrated in the judicial system of resolution of disputes in the field of taxation set out in the Portuguese constitution itself, as opposed to the investment tribunal in Achmea, set up on the basis of an international treaty. Yet, the Court failed to acknowledge that both in Achmea and in Ascendi, the choice of arbitration involved public authorities, which were then involved in the arbitral proceedings themselves: the jurisdiction of the investment tribunal in Achmea is based not only on the BIT, but also on the Dutch and Slovak statutes ratifying that BIT. By virtue of such ratification, the BIT itself can be considered part of the legal orders of both contracting parties. Secondly, the tribunal in Achmea was not less ‘permanent’ than the one in Ascendi. The investment tribunal indeed had permanent compulsory jurisdiction under article 8 of the relevant BIT and by the reference to the rules of the Stockholm Chamber of Commerce (SCC) and the UNCITRAL rules, which, just as in Ascendi, “define[d] and frame[d] the applicable procedural rules.” In relation to the compulsory jurisdiction of the arbitration tribunal, it is significant that in both Achmea and Ascendi the claimant could decide whether to bring its case before national courts or the arbitration tribunal. The analysis of the requisites of independence and impartiality of the arbitrators is probably the strongest point in the reasoning of the Court. While the independence and impartiality of some of the national courts of the Member States is far from unquestionable, the effectiveness of the rules on independence and impartiality of arbitrators is admittedly a highly debated issue even in the arbitration community itself. The reasoning of AG Wathelet also does not address the issue, in so that it only refers – not entirely correctly – to the fact that the Court had not questioned the independence and impartiality of the arbitrators in any previous cases. On the opposite, in Ascendi the Court did look at the appointment procedure of the arbitrators, although admittedly not in details. The AG rather unsatisfactorily reasoned only in the sense that the independence and impartiality of the arbitrators is guaranteed by the UNCITRAL Arbitration Rules, which inter alia lay down a procedure for challenging arbitrators where doubts exists in that sense. Be as it may, the Court did not base its distinction between the two cases on that point. All in all, after reading Achmea one is left with the idea that the Court was really making an exception to its previous case law on the possibility for arbitral tribunals to make preliminary references to the CJEU, rather than implementing it. An exception that appears to really be based on the application of a high degree of scrutiny on the compatibility of dispute settlement mechanisms sitting outside of international law with EU law, and particularly its system of judicial protection. That high degree of scrutiny also appears to emerge in the extremely broad language that the Court adopts in relation to the definition of the scope of the case, which seems to suggest the exclusion from the EU judicial system, and thus from the opportunity to refer preliminary questions, not only of the specific tribunal constituted under article 8 of the Netherlands-Slovak BIT, but also of any tribunal “such as that referred to in Article 8 of the BIT.”
Arbitral tribunals adjudicating on EU law matters – Eco Swiss
The ever-expanding scope of the powers and competences of the EU creates increasing jurisdictional overlaps between the system of judicial protection of the EU and arbitral tribunals. These overlaps have oftentimes been regarded by the CJEU as challenging its ‘monopoly’ to adjudicate on EU law conferred to it by article 344 TFEU, and as creating risks of awards conflicting with EU law. The situation is made more complex by the line of case law of the Court on the interpretation of what courts or tribunals of the Member States are for the purpose of the preliminary reference procedure under article 267 TFEU, which often excludes that arbitral tribunals may refer preliminary questions to the CJEU. A particularly representative example of this overlap is Eco Swiss, where the Court was faced with the case of an arbitral tribunal dealing with issues of competition law. As mentioned, Eco Swiss is a particularly interesting case in terms of understanding the approach of the Court to arbitration, among other things because the court resolved a dispute internal to the EU through an instrument typical of to international law, public policy. The case in Eco Swiss arose from a dispute on the termination of a licensing agreement concluded among the companies Benetton, Eco Swiss, and Bulova. The said agreement entailed a complex market sharing arrangement under which Eco Swiss was precluded from entering the Italian market and Bulova could not sell in all other Member States of the (then) EEC. The contract was governed by Dutch law and subjected disputes to arbitration. When Benetton gave its commercial partners notice of termination of the contract three years ahead of the contractual deadline, Eco Swiss and Bulova commenced arbitration against their contractual partner on the ground of early termination of the contract. The arbitral tribunal eventually found Benetton liable for damages. Benetton applied to the Dutch courts to resist to the enforcement of the award, invoking the ‘public order’ exception on the grounds that the terms of its own licensing agreement with Eco Swiss and Bulova violated EU competition law on the prohibition of certain market-sharing agreements among undertakings under article 85 EC (article 101 TFEU). As competition law rules were not considered rules of public policy in the Netherlands, the Dutch court deciding the case sought guidance from the CJEU on whether EU competition law could be regarded as a fundamental mandatory rule EU law capable of trumping the enforcement of an arbitral award. The Dutch court also asked the CJEU whether the parties could raise the issue of public policy for the first time during the national proceeding for the annulment of the award, while never having relied on it during the arbitration proceeding. The CJEU firstly observed that, in line with its previous decision in Nordsee, national courts had a duty to scrutinise the compatibility of arbitral awards with EU law and, when appropriate, refer issues to the Court, due to arbitral tribunal sitting outside of the EU legal system. The CJEU then moved to observe that competition rules are “fundamental” rules of EU law crucial for “the accomplishment of the tasks entrusted to the Community and, in particular, for the functioning of the internal market.” On this basis, the Court then endorsed the claimant’s argument that article 101 TFEU “falls among the national rules of public policy which national courts must examine when granting an application for annulment of an arbitration award” and that the “importance of such a provision led the framers of the Treaty to provide expressly, in [article 101 TFEU], that any agreements or decisions prohibited pursuant to that article are to be automatically void.” In so doing, the Court thus relied on the internal autonomy of EU law from the national law of the Netherlands – which renders immaterial that Dutch law does not include competition law rules in its public policy exception within the scope of EU law and mandates that national legal systems comply not only with the black letter of the law, but also effectively uphold its aims. While the fundamental principle of the decision of the Court to give prevalence to competition law rules over the award of the arbitral tribunal is certainly understandable from an EU law perspective, Eco Swiss is problematic in so that it once again upholds a high standard of compatibility of commercial awards with EU law.To begin with, it remains unclear why the CJEU resorted to the concept of ‘public policy’ rather than simply relying on its established doctrine of primacy of EU law. Admittedly, public policy is liable to sort effects that are very similar to primacy, in so that it preserves the full effectiveness of the municipal legal order (in this case, EU law) at the expenses of the rule of the ‘foreign’ legal system (in this case, Dutch law), which is rendered inapplicable. However, primacy of EU law has a very different scope compared to public policy: while the former entails a presumption of prevalence of the EU legal system over that of the Member States, the latter is an occasional exception to the application of foreign rules that offend the fundamental tenets of a legal order. Conflict of law rules, to which public policy belongs, are normally based on exceptional considerations of structural incompatibilities of external rules with the fundamental tenets of a legal system. Yet, the CJEU in Eco Swiss did not engage in any justifications for its exceptional application of EU rules over arbitration or of the structural incompatibilities that may have justified that conclusion. Eco Swiss thus seems to unnecessarily create the additional complication of introducing in the case law of the Court a new legal category of ‘EU public policy,’ whose scope and existence remains to date still debated. One may speculate that the Court deemed necessary to do so because it considered the relationship between EU law and arbitration as a matter external to EU law, and thus outside of the scope of application of the doctrine of primacy. Or perhaps its conclusion simply was the outcome of the formulation of the preliminary question by the referring Dutch court. Be as it may, the Court in Eco Swiss seems to unduly apply to the external relations of the EU, a standard of review, that of primacy, that can only pertain to its relations with national courts. From another standpoint, the scope of the ‘public policy exception’ in Eco Swiss also appears to be framed very broadly. The language of the Court seems to suggest that all violations of EU competition law can be relied upon to trump arbitration awards, in so implicitly raising the bar for its application very high. This became especially evident in a Genentech, where the claimants sought to rely on Eco Swiss in a set-aside proceeding connected to the payment by the claimant of royalties connected to a licence agreement, which had allegedly exposed it to additional costs as compared with its competitors. The Court on that occasion narrowed down the scope of its previous decision in Eco Swiss, affirming that not all breaches of competition law triggered the public policy exception. However, dismissing the attempt in that sense of Advocate General Wathelet, the Court regrettably did not entertain a discussion on how to identify which breaches of competition law do trigger a public policy exception. Finally, Eco Swiss opens the door to abuses of EU law. As it will also be discussed in relation to Achmea, the main criticism that can be moved to Eco Swiss is that the Court effectively allowed the claimant to rely on the illegality under EU law of the terms of a contract from which it had itself benefited to escape its obligations towards its contractual counterparties.
Violation of article 344 TFEU – Mox Plant
In Mox Plant the Court famously considered whether article 344 TFEU, conferring exclusive jurisdiction on the interpretation and application of EU law upon the Court, stands in the way of Member States resorting to an international arbitration under an international convention –UNCLOS – to regulate disputes inter se on matters partly falling under EU law. As a matter of general background, UNCLOS is an international convention concluded in 1982, and entered into force in 1994, which governs the use of oceans and their resources. Its 157 signatories include the EU and all its Member States. UNCLOS provides that jurisdiction on the interpretation or application of its provisions – or of any other international agreement linked to that convention – belongs to one of the dispute settlement mechanism provided under article 287 of the same text (i.e., International Court of Justice, ITLOS, or arbitral tribunals). Article 282 UNCLOS also provides that parties that have already agreed to another binding dispute settlement mechanism, may agree submit their dispute to that mechanism in lieu of those provided under article 287 UNCLOS. Under article 293 UNCLOS, the law applicable to disputes arising from that text comprises the text of UNCLOS itself and other applicable rules of international law. Participation of the EU in UNCLOS is made possible through annex XI to that convention, titled “Participation by international organisations.” For what is relevant here, the annex required the EU to deposit a declaration of competence specifying which convention rights it would exercise in lieu of its Member States. The ‘Declaration of Community competence’ deposited by the EU reads in the relevant part “[w]hen Community rules exist but are not affected, in particular in cases of Community provisions establishing only minimum standards, the Member States have competence, without prejudice to the competence of the Community to act in this field.” Annex XI of UNCLOS also provides that UNCLOS obligations prevail over the obligations arising out of the founding instrument of the international organisations party to it in the event of a conflict. The facts of the case in Mox Plant can be summarised as follows. Pursuant to article 287 UNCLOS, in 2001 Ireland started arbitral proceeding against the United Kingdom over the operations of the Mox Plant on a site at Sellafield, on the Irish sea, on the grounds that it was increasing pollution in the Irish Sea. The claim brought by Ireland included a number of indirect references to EU law. Pending arbitration before ITLOS, the Commission launched a separate action against Ireland for failure to fulfil its obligations under the EU treaties. According to the Commission, Ireland should have instituted those proceedings before the Court of Justice, and referral to the ITLOS tribunal violated the exclusive jurisdiction of the Court under article 344 TFEU and the corresponding provision in the European Atomic Energy Community (EURATOM) treaty. The Commission also maintained that Ireland had violated its duty of cooperation under articles 10 and 292 EC (4(3) TEU, 344 TFEU) by failing to inform and consult with the Commission over the institution of the ITLOS proceeding. Ireland objected that the obligations invoked against the UK were those arising under UNCLOS rather than those under the EU treaties, and thus they did not fall within the matters on which the CJEU enjoys jurisdiction. Considering the points of EU law arisen in the case, the seized arbitral tribunal suspended hearings until the delivery of the decision of the CJEU. The Court approached the case at issue from a standpoint of competences. It firstly affirmed that the external competence of the (at the time) Community to conclude agreements on the protection of the environment, qualifying it as ‘mixed’ between the EU and its Member States. The Court then affirmed that mixed international agreements of which the EU is part – including UNCLOS – form “an integral part of the Community legal order.” With a view of examining the extent to which EU competence had pre-empted the competence of the Member States in the context of UNCLOS, the Court went on to analyse the extent to which the EU had in fact exercised its external competence in matters of environmental protection. Referring to the ‘Declaration of Community competence’ submitted by the EU at the act of accession to UNCLOS, the Court held that EU rules had to exist in order for the transfer of shared competence to take place, even though it was not “necessary that those rules be affected.” Having considered that “matters covered by the provisions of the Convention relied on by Ireland before the Arbitral Tribunal are very largely regulated by Community measures” the Court upheld its jurisdiction on the dispute at issue. Reasoning on the grounds of the autonomy of EU law, the Court then moved on to affirm that that such jurisdiction should be qualified as exclusive. Under the second head of the judgement, the Court also found Ireland in breach of article 344 TFEU. Dismissing Ireland’s claim that EU law had only been relied upon indirectly, by renvoi, the Court decided that Ireland had “submitted instruments of Community law to the arbitral tribunal for purposes of their interpretation and application” to the case at issue. The CJEU thus concluded that the fact that “some of the measures in question come within the scope of the EC Treaty and others within the scope of the EAEC Treaty” created a manifest risk that the autonomy of EU law would be affected, Ireland had breached its obligations under the EU treaties. The decision in Mox Plant appears to once again adopt a markedly protectionist stance of the EU legal order on the part of the Court and shows a high degree of scrutiny in the compatibility of international law with EU law. The impression that one gains from reading the decision is that, once again, the Court comes close to transforming what should be an exception of the incompatibility between EU law and international law on the grounds of autonomy, into a default approach of primacy of EU law – which typically characterises its internal relations. To begin with, the Court’s reasoning on the allocation of competences seems hardly reconcilable with the wording of the ‘Declaration of Community competence.’ As seen above, that declaration set out that matters on which the EU has exercised its competences, but that are not affected by the actions of the Member States under UNCLOS, rest with Member States. In establishing its jurisdiction, the Court instead create what seems a new rule that Union competence is triggered even when a significant part of the dispute is deemed to relate to an area where pertinent EU instruments exist – regardless of whether those instruments are affected by the dispute. The Court took for itself the role to identify in those cases the parts of the dispute relating to the relevant international agreement and the parts which instead fell within its jurisdiction. Read together with the exclusivity of the Court’s jurisdiction, the decision leads to the conclusion that the general existence of EU instruments on matters that just partially fall within the scope of the subject matter of a dispute to which Member States are a party to is sufficient in itself to give to the Court the power to decide on its own jurisdiction at the exclusion of any other decision-making mechanisms. In so doing, the Court de facto ends up extending its own jurisdiction beyond the letter of the ‘Declaration of Community competence’, the wording of UNCLOS, and the jurisdiction conferred upon it by the EU treaties. The issue is not theoretical: given the broad scope of EU instruments in the field of marine environment, the practical effect of the decision virtually is that all international disputes between Member States connected to marine environmental protection will be considered as falling under the exclusive jurisdiction of the Court. The conclusion is even more staggering when read against the letter of annex XI UNCLOS, which provides for a general affirmation of primacy of the provisions of that agreement over the treaty obligations of the participating international organisations. The stance of the Court on the issue of the competences of the EU in relation to the dispute also appears troublesome in so that it risks depriving certain disputes of a resolution forum: if the threshold to trigger the EU’s exclusive jurisdiction to determine the scope of its competence is a significant part of the disputes, then the issue arises of what will happen to the parts of the dispute that, as a result of the Court’s decision, are regarded as not falling within its competences. The consequences of this determination are very relevant. In Mox Plant the Court did not address Ireland’s concerns relating to terrorism, finding that the issue was absorbed by the argument of the majority of the dispute. Arguably, that binds future claimants in similar disputes to go through the lengthy process of awaiting a decision of the Court on its jurisdiction before being able to bring a dispute on issues that the CJEU eventually will regard as falling outside of its jurisdiction. The second part of the decision of the Court, on the violation by Ireland of (what is today) article 344 TFEU, confirms the protectionist approach of the Court in Mox Plant. Firstly, the Court excluded that EU law could be considered by an international tribunal even indirectly or as a matter of fact as international courts and tribunal normally do when deciding disputes regulated by more than one set of rules. In so doing, the CJEU expressly rejected that Ireland could rely on the tool of renvoi, even if only for clarifying the legal issues involved. Interestingly, while admitting that the instrument of renvoi is “a frequently used juridical technique designed to guarantee the harmonious coexistence of rules deriving from different legal orders” the Court seems to adopt a blanket ban of its use in the international disputes of Member States. This approach of the Court is yet another evidence of the very high level of compatibility with EU law of the arbitral proceeding, as the CJEU turns an exception to the normal functioning of international law (i.e., to impossibility to use of renvoi and judicial comity) as a fully-fledged rule in the external relations of the EU.
It also seems relevant to note that the second head of the decision in Mox Plant also makes again use of the same ‘precautionary language’ discussed above in relation to the preliminary opinions under article 218(11) TFEU. Specifically, the Court held that the autonomy of the EU law would be upset by the submission to an arbitral tribunal of a dispute that “involves the risk that a judicial forum other than the Court will rule on the scope of obligations imposed on the Member States pursuant to Community law.” The Court’s decision in Mox Plant appears to narrowly escape a systemic clash of legal regimes almost ‘by accident.’ The Court ultimately recomposes the pieces of the different legal regimes involved of EU law and UNCLOS through a reference to article 282 of the UNCLOS, which provides – as an exception to the usual dispute settlement mechanisms in that convention – that parties to a dispute that have already agreed to another binding dispute settlement mechanism (in this case, the CJEU) may agree submit their dispute to that mechanism. It remains however doubtful that the Court’s decision would have been different in the absence of that clause: the reasoning in Mox Plant seems indeed to be heavily based on arguments of a general ‘hierarchical prevalence’ of EU law rules over international conventions rather than on an attempt to reconcile legal regimes. It is noteworthy that the broad reading of the Court of article 344 TFEU in Mox Plant had a direct impact on state-to-state arbitration among EU Member States, signalling a definitive end of this practice. A fate later shared by intra-EU ISDS arbitrations in Achmea.
Autonomy issues with ISDS arbitration
With the exception of Mox Plant, which regards state-to-state arbitrations based on the application of rules of public international law, the other cases discussed above consist of commercial arbitrations. With the entry into force of the treaty of Lisbon, which conferred to the Union competence in the field of foreign direct investments, the Court became directly concerned with the compatibility of the mechanism of investor-state arbitration with EU law. The three main cases decided to date by the CJEU on the interface between investment arbitration and EU law are Opinion 2/15, Achmea, and Opinion 1/17. In all three instances, the Court expressly recalls its established principle that the “competence of the European Union in the field of international relations and its capacity to conclude international agreements necessarily entail the power to submit to the decisions of a court which is created or designated by such agreements as regards the interpretation and application of their provisions.” However, at least in the first two of them, Achmea and Opinion 2/15, the Court in practice adopts its ‘typical’ very high degree of scrutiny in the compatibility with the autonomy of EU law in relation to investor-state arbitration mechanisms.
Opinion 2/15
On 16 May 2017, the Court published its Opinion 2/15 on the conclusion of the Free Trade Agreement between the European Union and the Republic of Singapore (EUSFTA). While that preliminary opinion of the Court deals with investment arbitration mainly from a standpoint of competence, the underlying reasoning and outcome of the decision are nevertheless relevant towards an analysis of the scrutiny of the Court of the ISDS mechanism included therein. In Opinion 2/15 the Court was called to rule on the allocation of competences between the EU and its Member States about the provisions included in the draft EUSFTA, including the ISDS mechanism set out in chapter 9, section B of the EUSFTA, on which the present analysis will focus. At the outset of the its decision, the Court firstly recalled the established principle in its case law that the EU possesses an implied power to set up dispute settlement mechanisms in its areas of competences. The Court then reiterated its longstanding case law that dispute settlement mechanisms are normally considered as ‘ancillary’ to the substantive provisions that they accompany. It follows from that that, in principle, the exclusive scope of the Union’s competence in the field of FDI should provide it with the exclusive competence to set up the dispute settlement mechanisms under chapter 9 of the EUSFTA. However, the Court then moved on to distinguish the case of ISDS from its previous jurisprudence. The CJEU observed that, after the ratification of the EUSFTA, the ISDS mechanism in chapter 9 would allow investors to submit the dispute to arbitration without Member State being able to oppose this. Chapter 9 would thus allow investors to circumvent national courts and sue Member States (and the EU) directly before an arbitral tribunal. For that reason, the Court concluded, ISDS mechanisms could no longer be considered ‘ancillary’ to the Union’s competence under article 207 TFEU “within the meaning of the case-law recalled in paragraph 276 of this opinion and cannot, therefore, be established without the Member States’ consent.” On the basis of that reasoning, the Court declared that the ISDS provisions in the EUSFTA fall within the mixed competences of the EU and its Member States and will thus have to be signed and concluded both by the European Union and by each of its Member States The reasoning of the Court seems unconvincing and evidences a high threshold of scrutiny in the compatibility with EU law of ISDS mechanisms for various reasons. To begin with, it is not self-evident on what legal grounds the Court deviates from the established principle of the ancillary nature of dispute settlement mechanisms to the subject matter to which they attain, which would have led to the conclusion that ISDS fall within the exclusive competence of the EU in the field of foreign direct investments. The reasoning of the Court that the creation of investment tribunals falls within the mixed competence of the EU and its Member States because, after the ratification of the EUSFTA, Member States could be faced with investment disputes without “being able to oppose” them, overlooks the essential tenet of investment arbitration that ‘consent’ to investment arbitration is considered given at the stage when states enter into an investment agreement. That should have been obvious from the outset, at the stage of the agreement of the Council mandate to negotiate to the Commission, that, were Member State and the EU to decide on whether to set up an ISDS mechanism. It remains unclear why, once this consent has been given, they should have the right to be consulted again, and possibly even revoke that consent. Should that right to a ‘double’ consultation be applied to the entire scope of the Union’s external relations, it would become impossible for the Commission to negotiate international agreements due to the continued changes of heart of the Member States. Secondly, Opinion 2/15 creates another unexplained contradiction with the established case law of the CJEU: the decision that certain EUSFTA chapters, including chapter 9, on which the EU and its Member States possess mixed competences must be ratified by both (mandatory mixity) contradicts the previous case law of the Court that mixed agreements can be concluded by the EU alone (facultative mixity). It is noteworthy that the position of the Court on mandatory mixity has a relevant impact on the external relations of the EU, as it results in an added procedural complexity in the conclusion of such agreements, that need to be ratified by the EU and all national Parliaments of the Member States. Even more strikingly, the derogation of the Court from its case law on mandatory mixity seems to be made ad hoc for ISDS: the Court went back to its established position on mixity as being facultative shortly after the publication of Opinion 2/15, in case C-600/14. Thirdly, the reasoning of the Court is less than satisfactory and inadequate to support the deviations from the established case law on the ancillary nature of dispute settlement mechanisms and mandatory mixity. The argument of the Court on the ancillary nature principle closely resemble an assumption: that that mechanism “remove[s] disputes from the jurisdiction of the courts of the Member States.” The CJEU does not attempt to further motivate its deviation from the case law on mandatory mixity. That course of action is particularly regrettable for three reasons. To begin with, because it introduces uncertainty in the case law of the Court. Further, because the composition of the Court, which sits in Full Court, affords to the decision special importance in terms of the precedent that it sets for future cases. A more elaborate reasoning would arguably have thus been in order. Finally, because of the nature as a preliminary opinion of the decision, which, as mentioned in chapter […] above, confer to the Court exceptional powers to impact on the EU policy in the external relations and needs thus to be carefully justified. Once again, what appears troublesome in the approach of the Court in Opinion 2/15 is not the exceptional importance that the Court attributes to the judicial system of protection of the EU as an essential element of the constitutional structure of the EU. It is in the mandate to of the Court to ensure that the main constitutional elements of EU law are safeguarded from damaging external interferences. It is rather the very high degree of scrutiny of the Court in the functioning of ISDS mechanisms within the framework of EU law – which transforms its power to exceptionally intervene to preserve certain constitutional traits of the EU legal system into a default approach of primacy of EU law over international law – that rather appears problematic.
As Mox Plant did for intra-EU arbitration, Opinion 2/15 may also well mean the end of ISDS mechanism in EU free trade agreements. As Sarmiento notes, the condition of their inclusion in a negotiating agreement to the approval of the national parliaments of the Member States in practice may well be enough to dissuade the negotiating parties from resorting to them. So that quite fitting seems the analogy made by the author of Opinion 2/15 with a “silver bullet” for ISDS mechanisms in the free trade agreements of the EU.
Achmea
In the context of a preliminary reference procedure, in Achmea the Court was called for the first time to assess the compatibility of ISDS mechanisms with the legal system of the EU. Achmea has rightly been regarded as having ‘constitutional’ importance in so that it adds an important tile to the mosaic of the external autonomy of the EU in relation to dispute settlement mechanisms: the Grand Chamber of the CJEU declared that articles 267 and 344 TFEU preclude the operativity of investment arbitration clauses “such as” that included in article 8 of the Dutch-Slovak BIT “under which an investor from one of those Member States may, in the event of a dispute concerning investments in the other Member State, bring proceedings against the latter Member State before an arbitral tribunal whose jurisdiction that Member State has undertaken to accept.” As an effect of the decision of the Court, the referring national court ‘set aside’ an investment award obtained by a Dutch investor against the Slovak Republic. The case in Achmea regarded an investment arbitration clause included in a bilateral investment agreement concluded between two Member States ‘intra-EU’ BIT. The ISDS provision in article 8 of the (at the time) 1992 Netherlands-Czechoslovakia BIT at issue in Achmea is one of the provisions that remained in force after the accession to the EU of the Slovak Republic in 2004. In 2008 a Dutch investor operating in the sickness insurance market (Achmea) brought arbitral proceedings against the Slovak Republic under that clause. Achmea argued before the investment tribunal constituted under the BIT that the reversal of the liberalisation of the sickness insurance market by the Slovak Republic had infringed its rights under the BIT to fair and equal treatment and to free transfer of payments. The proceeding was governed by the 1976 UNCITRAL arbitration rules and the seat of the arbitration was Frankfurt am Main. The applicable law clause, found in article 8(6) of the relevant BIT, provided that disputes under that article shall be decided on the basis of the law, taking into account in particular though not exclusively: the law in force of the Contracting Party concerned; the provisions of this Agreement, and other relevant agreements between the Contracting Parties; the provisions of special agreements relating to the investment; and the general principles of international law. The tribunal eventually found in favour of the Dutch investor and ordered the Slovak Republic EUR 22.1 million in damages. The Slovak Republic brought an action before the German courts to have the final award set aside. Particularly, the Slovak Government argued that the country’s accession to the EU had rendered the jurisdiction of the arbitral tribunal in article 8(2) of the BIT incompatible with EU law. In particular, the Slovak Republic argued that the entire arbitration agreement in the Dutch-Slovak BIT was null and void and contrary to public policy inter alia for violation of EU law, in the form of articles 18 (prohibition of discrimination on the grounds of nationality), 267 (preliminary reference) and 344 (exclusivity of the Court’s jurisdiction) TFEU. Although unconvinced by the Slovak Republic’s arguments, the German Federal Court of Justice decided nonetheless to seek guidance from the CJEU due to the novelty of the issues raised. Invested with that preliminary question, the CJEU eventually found the incompatibility of the arbitration clause included in the BIT with articles 267 and 344 TFEU. The Court deemed unnecessary to decide on the issue of non-discrimination under article 18 TFEU. The reasoning of the CJEU is divided in three main parts: (i) evaluation of whether the arbitral tribunal may be called to decide on the interpretation or application of EU law; (ii) if that this is the case, whether the arbitral tribunal is a court or tribunal of a Member State under article 267 TFEU able to make a reference to the CJEU; and (iii) should that conclusion be negative, is the arbitral award liable to be reviewed by the national courts of Member States. In the first part of its decision, the Court concluded that the disputes which the arbitral tribunal mentioned in article 8 of the Dutch-Slovak BIT is called to resolve are liable to relate to the interpretation or application of EU law. Considering that EU law is, by definition, an constituting part of the domestic legal order of the Member States, the Court reasoned that the broad wording of the applicable law clause in article 8(6) the relevant BIT – which notably allowed it to decide on the basis of the “law in force of the Contracting Party concerned” – meant that the tribunal “may be called on to interpret or indeed to apply EU law, particularly the provisions concerning the fundamental freedoms, including freedom of establishment and free movement of capital.” The Court thus concluded that such disputes should be decided by the EU judicial system. In the second part of its decision, the Court stated that an arbitral tribunal “such as that referred to in Article 8 of the BIT cannot be regarded as a court or tribunal of a Member State within the meaning of Article 267 TFEU.” This point was discussed in section […] above, in relation to the interpretation of the CJEU of the definition of ‘court or tribunal.’ For what is relevant here, the Court decided that the “exceptional nature” of investment tribunals does not allow them to be considered part of the judicial system of the either Member State involved (Netherlands or Slovak Republic). In the third part of its decision, the Court concluded that an arbitral award delivered by an arbitral tribunal “such as” the tribunal referred to in article 8 of the BIT is not subject to review by a court of a Member State. The reasoning of the Court mostly revolves around the flexibility intrinsic to the functioning of investment tribunals. The Court found that the discretion left to investment tribunals to determine their own rules of procedure – in particular, the seat and applicable law –, coupled with the limited grounds for review under the national law eventually selected – mostly public policy – entailed the possibility that these awards might escape judicial review by the national courts of the Member States. The Court acknowledged that, as affirmed in Eco Swiss, a limited review of arbitral award is in principle essential for arbitration to function efficiently. However, it then moved on to find that ISDS tribunals constitute an exception to that rule. The Court then moved to distinguish between commercial arbitration, at issue in Eco Swiss, from investment arbitration, the subject matter of Achmea, on the grounds that while the former “originates in the freely expressed wishes of the parties”, the latter derives “from a treaty by which Member States agree to remove from the jurisdiction of their own courts, and hence from the system of judicial remedies which the second subparagraph of Article 19(1) TEU requires them to establish in the fields covered by EU law disputes which may concern the application or interpretation of EU law.” The Court then concluded that the fact that the fact that mechanisms “such as” the one under article 8 of the BIT call into question the effectiveness of EU law, affect the preliminary ruling procedure provided for in article 267 TFEU and appear incompatible with the principle of sincere cooperation, make it incompatible with the “autonomy” of EU law. The decision in Achmea appears to share the same high threshold of scrutiny adopted by the Court in Opinion 1/15 on the EUSFTA and is questionable both in terms of style and reasoning. The weakness of the reasoning is particularly evident through the “double look” typical of the decisions of the Court, i.e., the contrast with the Opinion of Advocate General Wathelet, who notably arrived at opposite conclusions.
In terms of style of the decision, it is notable that the Court dismissed the compatibility of the longstanding institution of investor-state arbitration clauses in a mere 31 paragraphs. The brevity of the decision of the Court is particularly noticeable when compared to the 240 paragraphs of the opinion in the same case of the Advocate General Wathelet. The brevity of the judgement translates in little intelligibility of the decision of the Court. Among other things, that emerges from the puzzling distinction introduced by the CJEU between commercial and investment arbitration– two mechanisms that are rather similar in so that they are based on the consent of the parties and exist for the purpose of providing an alternative to court proceedings. What also stands out in terms of style is that the Court dedicates most part of its judgement to the re-statement of established constitutional principles of EU law and its own previous case law and little attempt is made to engage with the substance of the facts of the case. Finally, the language adopted by the CJEU also strikingly recalls the ‘precautionary’ stance taken in relation to preliminary opinions under article 218(11) TFEU. The Court observes that “the arbitral tribunal referred to in Article 8 of the BIT may be called on to interpret or indeed to apply EU law”, that “disputes falling within the jurisdiction of the arbitral tribunal referred to in Article 8 of the BIT may relate to the interpretation both of that agreement and of EU law,” “an investor from one of those Member States may, in the event of a dispute concerning investments in the other Member State, bring proceedings against the latter Member State before an arbitral tribunal whose jurisdiction that Member State has undertaken to accept,” “by concluding the BIT, the Member States parties to it established a mechanism for settling disputes between an investor and a Member State which could prevent those disputes from being resolved in a manner that ensures the full effectiveness of EU law, even though they might concern the interpretation or application of that law.” The result is a judgement that seems closer to an interpretative declaration of principle of what EU law is, rather than a critical application of existing rules to a set of controverted facts. The decision is difficult to understand also in terms of substance, from both an international and EU law perspectives. From the first standpoint, on the grounds of autonomy, in Achmea the CJEU quietly moves away from the general principle of international law of pacta sunt servanda, which itself had considered as having a ‘fundamental’ nature in the EU. By issuing a declaration of incompatibility of prior Member States’ international commitments with EU law, the Court allowed posterior EU law to prevail over them, in so retroactively undermining the rights of investors entailed in those commitments and without even mentioning the legal basis for its decision. Notably, the Court disregarded its “well-established case-law” that, decisions on issues on the interface of EU law and international law should be based on general customary international law on treaty interpretation expressed in article 31 of VCLT and Article 31 of the 1986 Vienna Convention on the Law of Treaties between States and International Organisations or between International Organisations; and that the interpretation most in line with international law should be privileged. The CJEU rather seems to assume that EU law has replaced the previous international obligations of the Member States and simply ‘wishes away’ the previous obligations of the Member States. Nor Achmea clarifies the legal fate of the investment arbitration clauses incompatible with EU law, i.e., whether they are ‘terminated’, whether a succession of treaties has taken place at the act of accession of the relevant Member States, and, if so, on which bases. From a standpoint of EU law, the reasoning of the Court in Achmea sits uncomfortably with article 351(1) TFEU – which provides that the rights and obligations arising from the international agreements of the Member States concluded before the accession shall not be affected by the provisions of the EU treaties. It is true that the language of the provision, which refers to “Member States on the one hand, and one or more third countries on the other,” as interpreted by the CJEU would seem to refer to compliance of obligations of Member States vis-à-vis third countries, thus with the exclusion of the obligations that individuals may derive from the relevant international treaty. However, that interpretation appears excessively limiting when it comes to protecting certain fundamental individuals rights conferred by the pre-existing international treaty in conflict with EU law, such as those included in investment treaties, as the Court itself has acknowledged in Burgoa and Commission v Portugal. In those cases, the Court indeed overcame the restrictive interpretation of article 351 TFEU as establishing a strict dichotomy between a third state’s rights and the Member State’s obligation towards that third state. Specifically, by affirming that “Article 234 [351 TFEU] of the Treaty does not have the effect […] of adversely affecting the rights which individuals may derive from such an agreement” the Court expressly opened to the possibility of the ‘obligations’ in article 351(1) TFEU being owed to residents or companies operating in the Member States. Once it is acknowledged that article 351(1) TFEU does require Member States to comply with their international obligations towards individuals, the ‘automatic termination’ of international treaties conferring rights on private persons that the Court appears to endorse in Achmea conflicts with the Court’s interpretation of the second comma of the same provision: the previous case law of the Court on article 351 TFEU foresees that incompatibilities between the EU treaties and international agreements of the Member States that cannot be overcome may lead at the most to an obligation for the Member State concerned to denounce the treaty, but do not authorize incompliance. Achmea also is problematic from a standpoint of EU law in so that the Court did not engage in an interpretation of the Dutch-Slovak BIT to overcome the “incompatibilities” with EU law under article 351(2) TFEU despite having previously found that an obligation in that sense exists on the national courts of the Member States. Notably, in Budějovický Budvar the Court found that national courts must minimize possible incompatibilities between the EU treaties and the international agreements of the Member States “to the extent possible and in compliance with international law, in such a way that it is consistent with Community law.” Numerous parts of the decision of the Court in Achmea further evidence a particularly high degree of scrutiny of the compatibility of investment arbitration with EU law. To begin with, the Court seems to treat the international agreement at issue as a purely internal constitutional matter, to be resolved only on the basis of internal rules, without considering its international nature and broader ramifications on the external relations of the EU and its Member States. It is significative that the CJEU relied in its judgement in Achmea on Associação Sindical dos Juízes Portugueses, a defining decision in the internal relations which defined the procedural and institutional safeguards that allow the courts of the Member States to carry out an effective judicial review. The ‘domestic law’ approach taken by the Court in Achmea was confirmed by the President of the Court himself some months after the decision during a conference in London. On that occasion, President Lenaerts observed that the judgement stands on the ground of mutual trust, which stands in the way of Member States voluntarily subtracting entire segments of jurisdiction from their respective national courts. Yet, the conclusion that the legal issues raised by Achmea are ‘internal’ to EU law is neither self-evident from the text of the judgement nor it necessarily follows from the facts and consequences of the case. Admittedly, the subject matter of the case is an international agreement with internal effects between two Member States pre-dating the accession of one of them (the Czech Republic) to the EU. However, it is a fact that Achmea is not a purely internal case, as the Court seems to regard it. Firstly, the source of the Member States’ obligation remains an international agreement still in force. Further, the effects of the Dutch-Slovak agreement create rights on individuals under international law, which the Court overlooks in its decision. In so doing, as Gourgourinis also notes, the Court almost shifts the ‘burden of proof’ on the compatibility between EU law and article 8 of the Dutch-Slovak BIT over the proponents of that compatibility. From yet another standpoint, the outcome of the decision of the Court in Achmea also seems so to extend of the jurisdiction of the Court beyond the confines of EU law. To begin with, analogously to the decision in Mox Plant, Achmea also expressly rejects that EU law may be considered by any court or tribunal sitting outside of the EU even as a ‘matter of fact’ when adjudicating on the dispute. In so doing, the Court confirms its blanket ban on the instrument of renvoi, an instrument that was observed to be widely used in international law, and that any dispute involving the interpretation or application of EU law, even if only indirectly or by renvoi, will need to be decided within the system of judicial protection of the EU. Further, by affirming the jurisdiction of the EU judicial system over international standards of protection, such as the ones included in the Netherlands-Slovakia BIT, the CJEU effectively turns them into a matter of EU law, even if their scope and tools of protection remain partially different: arguably, substantive standards of protection set out in investment treaties, such as FET clauses, or non-discrimination, are liable to different interpretation in the EU and in international law. In addition to this, even where the scope of the substantive standards of protection included in investment agreements could be deemed to overlap with those set out under EU law, it is undeniable that the procedural protection offered under the BIT, the ISDS mechanism, has a different scope compared to the system of judicial protection of the EU.
The reasoning of the Court on the point that that national courts and tribunals may not be able to review arbitral awards also appears to evidence a very high degree of scrutiny in the work of investment tribunals. To begin with, the reasoning of the Court relating to the difficulty that arbitral tribunal’s decisions may only exceptionally be scrutinised by the national courts of the Member States almost paradoxical – considering that in Achmea the Court is called to decide on a preliminary reference from a German court. Further, the reality of investment arbitration is that awards against states will normally go through a process of enforcement before national courts. It is true that, unlike the New York Convention, awards rendered in the framework of ICSID convention can be reviewed by national courts only on very limited grounds. Thus, ICSID awards have the potential to remain outside of the system of judicial protection set out under article 19 TEU. However, Achmea was administered under the UNCITRAL rules and not under the ICSID convention. It is also debatable that the Court intended to make that difference: the CJEU in Achmea did not make mention of the differences between the two regimes of the ICSID Convention and the New York Convention and their effects in the two cases the compatibility with EU law and there is no evidence that the different function of the two regimes may have impacted on its final decision. The Court even appears rather unsatisfied with the fact that national courts and tribunal may set aside arbitral awards on the ground of public policy, which it qualifies as “limited.” A possible reading of this argument is that the Court did not intend to generally play down the importance of public policy as a conflict of law rule, and that it is simply repeating the wording used by the national court in formulating the preliminary reference, without taking a specific stance on the issue. The referring German court indeed interrogated the Court on the compatibility with EU law of a review of the arbitral award which could “validly be limited solely to breaches of fundamental provisions of EU law.” Yet, where one was instead to take the Court literally, and hold that ‘public policy’ is not a safe enough safety valve in the relationship between international arbitration and EU law, then the decision in Achmea would effectively heighten the already high standard of compatibility of EU law with international law set out in previous decisions: by that token, any interference, however small, of international law with of the EU legal system could become an intolerable intrusion in the autonomy of EU law. The implications of Achmea are further extended by the language used by the CJEU: concluding that provisions “such as Article 8” of the Dutch-Slovak BIT are incompatible with articles 267 and 344 TFEU, the Court left the door open to speculations on the scope of the judgement, and in particular whether it applies to all intra-EU BITs, the Energy Charter Treaty, and even to the BITs of the Member States with third countries. As it became particularly evident in the three different interpretative declarations that Member States adopted some months later on the scope of Achmea, that approach of the Court leaves much scope for uncertainty and speculations on the broad scope of ‘external autonomy,’ similarly to what Eco Swiss had done in relation to the concept of ‘public policy.’
Opinion 1/17
On 7 September 2017, Belgium submitted to the Court a request for an opinion pursuant to article 218(11) TFEU, relating to the compatibility with EU law of the ISDS dispute settlement mechanism included under section F of chapter eight (Investment) of the CETA, the ICS. For what is relevant here, in Opinion 1/17, Belgium requested the Court to review the compatibility of the CETA tribunal with the autonomy of EU law on two main grounds: in relation to the exclusive jurisdiction of the Court under article 344 TFEU, and as regards to the safeguard of the effectiveness of EU law from a standpoint of protection of the right to regulate. Little over a year after Achmea, the Court ruled on the compatibility of the CETA tribunal with EU law. Following the standard format for preliminary opinions under article 218(11) TFEU, the Court started its analysis with a re-statement of the constitutional principles withstanding its decision. As in Achmea, the Court recalled the essential constitutional nature of EU law and the principle of autonomy of the EU legal system, and made reference to the principled compatibility with EU law of international dispute settlement mechanisms included in the international treaties of the EU. However, to that list the Court added for the first time the acknowledgment of the reciprocal nature of the international agreements of the EU, i.e., the need for the contracting parties to create equivalent substantive and procedural standards of protection that apply to both. The Court tied this principle in with the overarching treaty policy aim of “the need to maintain the powers of the Union in international relations.” The reciprocal character of international agreements, the Court went on to explain, has two important consequences: on the one hand, it implies that the provisions of such international agreements may be interpreted by “courts and tribunals of the non-Member States with which those agreements were concluded or that of the international courts or tribunals that are established by such agreements.” On the other hand, it allows that the EU may confer jurisdiction to international courts or tribunals established under such agreements, such as the CETA tribunal, provided that such courts or tribunals do not interpret or apply EU law and upset the constitutional allocation of powers in the Union. Following the opinion of Advocate General Bot on that point, the Court then acknowledged that an assessment of whether the ICS adversely affects the principle of autonomy of EU law needs to be framed in the context of an “analysis of the reciprocal aspect of the desired substantive and procedural protection.” The Court then moved on to clarify that the dispute settlement mechanism in CETA is compatible with EU law because (i) it is not liable to interpret or apply EU law; and, (ii) EU institutions are not prevented from operating within EU constitutional framework by its decosopms. In relation to the first point, the Court declared itself satisfied that certain safeguards included in the CETA agreement are enough to avoid any meaningful interference with the system of judicial protection of the EU. The Court referred to the provision on applicable law, article 8.31.2 CETA, which provides that the law applicable to the dispute is the international law and that the tribunal may to take into account “as appropriate, the domestic law of a Party as a matter of fact.” The CJEU also recalled article 8.28.2(b) CETA, relating to the jurisdiction of the appellate body to correct “manifest errors in the appreciation of the facts, including the appreciation of relevant domestic law” which it however regarded as “clear[ly]” not being amenable of an interpretation in the sense of conferring on the ‘Appellate Tribunal’ of the ICS jurisdiction to interpret domestic law. Equally, the Court did not regard the jurisdiction of that Appellate Tribunal to review the appreciation of domestic law on the part of the first-tier tribunal as an issue with its compatibility with EU law. The Court also rejected the second argument formulated by Belgium, that the decisions of the CETA tribunals prevent the EU Institutions from operating within the constitutional framework of the EU treaties. For the Court, CETA contains sufficient safeguards to ensure that the assessment of primary legislation by its tribunals against the standards of non-discriminatory treatment (national treatment, most favoured nation) and investment protection (regulatory standards) would not in practice impact of the right of the Institutions to set the degree of protection of the public interest in the EU. Notably, the Court satisfied itself with the existence of procedural guarantees for the good operation of the EU Institutions (i.e., that the dispute settlement mechanism in CETA sits outside of the EU) without concerning itself too much with the potential practical impact of the decisions of the CETA tribunal (i.e., the consequences of decisions that impinge in the Union’s right to set its own standards of protection). On these grounds, the Court concluded that the CETA Tribunal does not adversely affect the autonomy of the EU legal order. Opinion 1/17 shows a move of the Court towards a reasonably low scrutiny of review of international agreements, in line with both the EU treaties and EU action in the international sphere. That emerges already in the recognition by the Court of the reciprocal nature of the international agreements of the EU, which so uncomfortably sits with Achmea and Opinion 2/13, that paves the way to a more conciliatory reading of external autonomy of the EU. It is also notable that in Opinion 1/17 the Court finally abandons the language of ‘possibility’ that characterised its earlier opinions under article 218(11) TFEU, showing a move towards a more restrained approach towards those instruments envisaged in section […] above. Moving to a detailed analysis of the decision, it is notable that the Court accepted that some interpretation of EU law on the part of the CETA tribunal is necessarily for its existence and functioning. That conciliatory interpretation was indeed expressly excluded in relation to other dispute settlement mechanisms in previous opinions, such as Mox Plant and Achmea. Yet, the Court did not expressly disowned that previous case law, but rather sought to distinguish it on the grounds of the existence of the express provision, in article 8.31 CETA, that the jurisdiction of the tribunal may take EU law into account as a “matter of fact.” That formalistic approach of the Court to the potential jurisdiction by renvoi of the CETA tribunal is unconvincing. Article 8.31 CETA merely spells out a rule of interpretation that is implied in the settlement of any international dispute – including disputes before the ITLOS tribunal at issue in Mox Plant, the European Court of Human Rights at issue in Opinion 2/13, and investment tribunals at issue in Achmea discussed above – that arbitral tribunals cannot directly adjudicate on issues falling outside of their jurisdiction, but may only do so indirectly to solve disputes before them. The idea is that arbitral tribunals will oftentimes have to refer to national laws and rules to interpret the facts of the case they are called to adjudicate upon and apply the law to them. However, the limited scope of their jurisdiction confines such interpretation and application to the extent necessary to solve disputes, so that the interpretation that they make opt for will not be binding on the domestic courts and institutions of the jurisdiction to which those rules pertain. It is thus unlikely that that interpretation will have a meaningful impact on the relevant legal system, and unthinkable that it may bind national courts, especially in a dualistic legal system as the EU. While the decision of the Court was most likely driven by deference towards the policy approach of the EU to ISDS reform rather than by the wording of article 8.31 CETA, the abandonment by the CJEU of the blanket ban on the jurisdiction by renvoi on EU law of arbitral tribunals represents a welcome advancement of the case law on external autonomy in the sense of a lowering of its degree of scrutiny of investment tribunals. It is also notable that the Court ditches its criticism previously expressed towards ISDS in Opinion 2/15 and Achmea that the possibility that investors may choose between bringing the dispute before the national courts of the Member States or investment tribunals represents a challenge to the autonomy of EU law.
A lower degree of review of the ICS tribunal with EU law is also evident in that the second part of the opinion, on whether the awards of the CETA tribunals may curtail the right to regulate the EU Institutions. Notably, the Court quickly considered itself satisfied that the CETA tribunal will not impact on the institutional balance set out in the treaties due to its lack of “jurisdiction to interpret the rules of EU law other than the provisions of the CETA.” Indeed, arbitral tribunals are routinely called to decide on the grounds of their constituent instrument and only indirectly interpret domestic rules. While it is true that the decisions of the CETA tribunal might influence the right to regulate of the EU – for instance, if adopted under ICSID rules, which will make them not be reviewable by the national courts of the Member States – it should be remembered that is part of the nature itself of international dispute settlement mechanism to influence the development of international relations. A ban on any potential interference with the right to regulate of international actors would thus translate in an exclusion tout court of the right of the Institutions to agree on international dispute settlement mechanisms or, for that matter, enter into any international agreements. For all these reasons, Opinion 1/17 is a welcome step ahead in the case law of the Court on the external autonomy of EU law, which complies with the standards of the relationship between EU law and international tribunals exercising outer scrutiny on the EU and the policy choices of the Council and the Commission. Following the observation of AG Bot that autonomy cannot mean autarchy and that the external relations of the EU are based on a concept of ‘reciprocity’, the Court carries out an overall reasoned assessment of the text of the CETA without jumping to conclusions on how that text may or may not be hypothetically be interpreted in practice. The one criticism that can still be moved to Opinion 1/17 is perhaps that it does not signal a clear enough break with the previous case law of the Court where autonomy instead resembles autarchy. By seemingly seeking continuity with its previous case law, the Court once again normalises the EU ‘exceptionalism’ in the international relations instead of taking the opportunity to clearly reframe the discussion on external autonomy in favour of a position of default compatibility with international law.
CHAPTER V – The issue of autonomy as primacy
1.introduction
Chapter […] analysed several decisions of the CJEU all broadly revolving around the topic of the constitutional interactions between EU and its system of judicial protection with outer scrutiny of international courts and tribunals. This chapter now moves to discuss the significance of that case law in the light of the constitutional evolution of international law in the treaties. The chapter is structured as follows. Section […] compares the high degree of scrutiny adopted by the Court in relation to international dispute settlement mechanisms to the approach of ‘primacy’ of EU law over national law. Section […] evidences the issues arising out of an approach to autonomy as ‘primacy’: in addition to contradicting the position of openness and dialogue towards engagement with international courts and tribunals set out in the EU treaties and entrenched in the external policy of the EU, that approach disregards the very different rationale of the internal and external relations, and also raises various concerns in terms of institutional balance, legitimacy, and overall role of the EU in the external relations. Section […] focuses specifically of the constitutional and policy issues raised by the approach of the CJEU to autonomy as primacy in relation to ISDS arbitration. Finally, some brief conclusions follow in section […].
2. External autonomy to justify an affirmation of primacy?
The case law on autonomy and outer scrutiny analysed above seems to show that, on the surface, the CJEU has accepted the approach of openness towards the engagement of the EU with international courts and tribunals entrenched in the EU treaties, as an element of the identity itself of the EU, and in the external policy of the EU. It is the settled case law of the CJEU that “an international agreement providing for the creation of a court responsible for the interpretation of its provisions and whose decisions are binding on the European Union, is, in principle, compatible with EU law” and that “the competence of the European Union in the field of international relations and its capacity to conclude international agreements necessarily entail the power to submit to the decisions of a court that is created or designated by such agreements as regards the interpretation and application of their provisions.” However, the CJEU has in practice often consistently returned to its approach of ‘safeguard’ of the origin on the seeming assumption that even merely potential encroachments on the defining traits of the EU legal system on the part of other international dispute settlement mechanisms, particularly judicial protection, could irreversibly compromise the autonomy of the legal system. That became, for instance, evident in the analysis of the preliminary opinions under article 218(11) TFEU, where the CJEU has with few exceptions excluded the compatibility of international dispute settlement system envisaged in international agreements of the EU with its interpretation of the external autonomy of the EU. As Opinion 2/13 demonstrates, that approach of safeguard of the Court even seems to strengthen in time. A quick survey of the existing literature seems to confirm that that approach of protection tends to be applied by the Court in relation to international law as a whole: while the Court has ‘borrowed’ or declared allegiance to general principles of international law and customary international law, their interpretation and application has often been ‘re-tailored’ to the specific characteristics of the EU. As Kuyper ironically observes, in all these decisions the Court attempts to base “the exceptional character of the Community legal order on normal rules of treaty interpretation.” For instance, it was evidenced how the principle of direct effect of EU law developed by the Court in Van Gend en Loos is not specific to the EU legal system but originates in international law as a tool to allow international rules to create some legal rights in national legal systems without first being translated into national rules. The case law of the Court relating to the rules of the VCLT, which codifies numerous international rules that are considered ‘customs’ in international law, is also particularly representative of how these rules were often reinterpreted to fit the specific paradigms of the EU. An interesting manifestation of this approach is the Court’s decision in Racke. The case arose from the challenge brought by a German importer of Yugoslav wines, Mr Racke, against the unilateral suspension by the (at the time) EC Council of a free trade agreement concluded with the Socialist Federal Republic of Yugoslavia. The suspension was motivated by the Council on the grounds of the outbreak of the hostilities in the region. In response to the preliminary question referred by the German Bundesfinanzhof on the validity of the suspension of the trade concessions arising under the agreement, the CJEU adopted a very loose interpretation of the international law doctrine of rebus sic stantibus: after suggesting that the applicant had relied himself on that principle (which was not the case), the Court went on to construe the doctrine of rebus sic stantibis as a rule of public international law, rather than an exception, as it is normally framed under international law, the proof of which was to be offered by the claimant. As the claimant had not offered such proof, the Court then moved on to evaluate whether the Council decision was based on a manifest error of assessment. That Court found that that was not the case. Thus, as Odermatt observes, while the use by the CJEU of the VCLT may at first sight seem to demonstrate fidelity towards international law, “one finds that the Court applies these rules in a fashion that often deviates from the way in which other courts and bodies have applied the same principles.” It was argued that the possibility that the EU may occasionally adopt a stance of ‘protection’ of the EU legal system in relation to the outer scrutiny of international courts and tribunals is arguably not in itself problematic: the autonomous constitutional identity of the EU confers upon its Institution “sovereign powers” to make determinations in relation to their relationship with international law and international bodies. That necessarily entails their right – and even duty – to avoid undue interferences of foreign law and foreign entities with the fundamental tenets of the system, as that could ultimately undermine its capacity to live up to its ideological blueprint of creating the “ever closer Union.” That conclusion finds support also in international law, where it is acknowledged that the possibility to decide the effects of foreign law and foreign relations in their territory is an essential part of state sovereignty, the “ultimate power” on the land. While that power finds its expression for the executive in policy decisions (e.g., to enter into a treaty, set up a court, or adopt a trade sanction), domestic judiciaries exercise that ‘sovereignty’ by means of mechanisms of conflict of law rules, through which they can limit, and even unilaterally reject, the application of foreign law in domestic law on the ground that its content seriously violates basic constitutional features of the jurisdiction in which its application is sought, “fundamental principle of justice, some prevalent conception of good morals, some deep-rooted tradition of the common weal.” For this reason, it appears not to be an issue per se that the CJEU relies on the doctrine of external autonomy of EU law to prevent or resolve three set of issues that in international law are normally dealt with by domestic courts through conflict of law rules:
Reject the choice of law made by its own rules. This is for instance the case of Opinion 2/13, where the Court expressly contradicts the EU treaty requirement of the EU accession to the European Convention on Human Rights in article 6(2) TEU based on its own determination of the potential damaging effects of that accession.
Reject foreign laws and decisions interfering with the forum’s morality and decency. These are for instance the cases deciding the jurisdiction on the interpretation and application of fundamental rights in the EU, again exemplified in a case such as Opinion 2/13, or the CJEU decision in Kadi, where the Court annulled a Council regulation implementing an UNSC resolution on the ground that the procedure for the adoption of the said regulation had breached the fundamental rights of the appellants. In these two cases, the CJEU preserves the autonomy of the EU legal system, in the form of the safeguard of ‘EU standards’ of protection of fundamental rights from interferences of outer bodies.
Prevent ‘injustice’ in judicial proceedings. Almost all the cases analysed in the previous chapter directly or indirectly aim to avoid outer interferences with the EU system of judicial protection, particularly with articles 267 and 344 TFEU. As mentioned, not only the CJEU regards the safeguard of its own system of judicial protection as an essential element of the constitutional structure of the EU, but also it seems to work on the assumption of its ‘qualitative’ superiority over other international dispute settlement mechanisms, particularly arbitration tribunals.
What instead seems problematic in the cases analysed above rather is the way in which the Court has been exercising its decision-making powers to protect the EU legal system from the interferences of the outer scrutiny of arbitration tribunals. Vis-à-vis a constitutional charter and that arguably makes of outer an element of the identity of the EU itself and a clear policy of favor of the EU towards participation in international dispute settlement mechanisms, the CJEU has often adopted an unreasonably high degree of scrutiny in the interferences with EU law that it is prepared to accept from international bodies, especially arbitral tribunals. In the analysed case law that high threshold of scrutiny emerged for instance in the fact that the CJEU often relies on the constitutional importance of the judicial system of the EU as a reason to declare even merely potential or indirect interferences of international dispute resolution mechanisms or their decisions incompatible with the autonomy of EU law. Exemplary are the preliminary opinions of the Court under article 218(11) TFEU on the compatibility with EU law of arbitral tribunals: it was seen above that in most instances the Court rejected such compatibility, even where no real threat to primary considerations of the EU existed. The inconsistent case law of the Court on the definition of court or tribunal under article 267 TFEU also arguably is evidence of a Court that remains wary of relationships with arbitral tribunals. That approach is hardly reconcilable with the EU treaties and the eternal policy position of the EU. The mentioned decisions of the CJEU in Mox Plant and Achmea have the ultimate effect of ‘liberating’ the Member States of their legal commitments under pre-existing international law agreements, i.e., the UNCLOS convention and the Dutch-Slovak BIT, in the absence of an effective threat for the EU legal order. More generally, when the CJEU diverges from the established international law approach to a certain issue, not only does it not “contribute to the strict observance and development of international law, but undermines it [and] may also contribute to the fragmentation of international law.” The case law of the Court discussed in the previous chapter also is hardly reconcilable with a view that ‘decentralization’ of powers and creation of effective outer control is part of the DNA itself of the Union since its inception and an element of its autonomy. Notably, in Opinion 2/13 the Court expressly provides that its exclusive jurisdiction on EU law “precludes any prior or subsequent external control” over the functioning of the system of judicial protection of the EU. That approach risks setting the Court and the EU on a trajectory of collision with the network of international institutions and dispute resolution mechanisms that ultimately is shaping the international law framework in which the EU and its Member States operate and could be used to both reinforce the identity of the EU and export its values and principles abroad. Nor can the very high degree of scrutiny adopted by the Court in the analysed decisions be justified under international law on the basis of the sovereign powers of the EU. To begin with, those powers are conferred to be exercised in the framework of the EU treaties. They thus cannot be used to justify the interpretation or application of the law that is in conflict with that framework. Further, international law remains based on the general rule of pacta sunt servanda, the principle of good faith, and the rule of comity among the members of the international community. That is arguably intrinsic to the nature itself of international law and international relations, which exist and develop among peers and are based on the collaboration of equally ranking entities with equal rights and duties. In an international law environment characterised by an increasing number of transnational relationships among legal and natural entities, it is thus in the common interest of the members of the international community that any derogations to common rules and unilateral withdrawal or limitations to reciprocal obligations are used sparingly. A low threshold of review of international rules applicable to the EU and of the operation of equally ordinated international actors thus remains the standard normally accepted in international law. Arguably, the adoption of a high degree of scrutiny by the Court in relation to the outer scrutiny of international courts and tribunals transforms external autonomy from a legitimate tool to justify the exceptional deviation of the EU from international law in order to safeguard its fundamental traits, into a full rule of primacy of EU law. The quasi-presumption of prevalence of EU law over international rules adopted by the Court seems indeed closer to the approach that the Court adopts in relation to national law and to its relationship with national courts rather than to an exceptional limitation to the operation of international rules to avoid irreparably fragmenting foreign interferences with core principles of the law or the land. Interestingly enough, it was seen above that the first reference to ‘autonomy’ in the case law of the CJEU appears in the French version of Costa, relating to the internal relations. The use of autonomy as primacy finds an interesting manifestation in relation to the development by the Court of the unclear notion of ‘EU public policy’ in Eco Swiss: it was seen on that occasion that, the Court confusingly failed to rely on the established principle of primacy to safeguard the application of the competition law rules from the effects of a decision issued by a decision making body sitting outside of the system of judicial protection of the EU. It resorted instead to the new notion of ‘EU public policy,’ a concept borrowed by international law to ensure prevalence of the EU rule above the international law one. Further, in Opinion 2/13 the CJEU also disregarded the “core value system” of human rights protection of the ECHR – as entrenched in the EU treaties – in a view to protect the internal characteristics of the Union’s legal system. The approach of the Court to autonomy as primacy seems to emerge in two main ways. Firstly, in certain cases, the Court seems to use autonomy as a tool to expand the scope of EU law (and of its own jurisdiction) at the expenses of the jurisdiction and operation of other dispute settlement mechanisms, and thus derive from that the primacy of EU rules over international law ones. The rule of thumb followed by the Court in the analysed decisions indeed appears to be that, once a link – however tenuous – between the subject matter of the jurisdiction of the court or tribunal and EU law has been established, the role of any other dispute settlement mechanism will be either extremely limited or even incompatible with EU law. For instance, it was seen above that the extremely broad interpretation of article 344 TFEU adopted by the Court in Opinion 2/13 almost equates to tout court excluding that the EU Institutions may enter in any negotiations setting up an international dispute settlement mechanism: as also noted by AG Kokott in her opinion in that case, a different interpretation could arguably lead to the paradoxical conclusion that the Court otherwise may end up ruling out the compatibility with EU law of longstanding international agreements, including the WTO agreement. By adopting a high threshold of compatibility of international court and tribunals and their decisions with the EU legal system, the Court arguably misreads the EU treaty rule and policy approach of principled engagement with foreign courts and tribunals, transforming those rules into the exception. That was also evidenced in the interpretation of the Court of the ‘Declaration of Community competence’ filed by the Union at the act of the accession to UNCLOS in Mox Plant, which had the effect of greatly extending the jurisdiction of the Court on disputes under that instruments. The Court held on that occasion that Union competence is triggered where a significant part of the dispute relates to an area where pertinent EU instruments exist – even where those instruments are not affected by the dispute. Notably, the interpretation of the significance of the part of the dispute rests solely with the CJEU. That use of autonomy as a tool to expand the jurisdiction of the Court is also evident in that the CJEU seems often to treat international law issues on which international courts and tribunals are adjudicating as purely domestic matters, subject to EU law, disregarding their broader impacts on international law. For instance, it is significant that the CJEU relied in its judgement in Achmea on Associação Sindical dos Juízes Portugueses, a defining decision in the internal relations that, it was seen above, defined the terms of a new, enhanced primary law right to judicial protection whose terms are only applicable to the internal constitutional relations of the EU between the CJEU and national courts. In Achmea the Court also seemingly overlooked the broader international law framework of the decision, nor did it take into account how that decision could fit in the broader picture of customary international law, comity in respect to concluded investment arbitration proceedings on the same issue, and, more broadly, of legal certainty and judicial protection of the investors involved.
Secondly, in the examined cases, the CJEU shows a tendency to use external autonomy to ‘Europeanise’ international law, demanding that other courts and tribunals conform to EU law in their operation and decision-making activity. Admittedly, ‘Europeanisation’ of international law is to an extent the inevitable result of both the growing unification of EU law and the increasing expansion of its competences. On the one hand, a strong normative power to impose its values and standards over third countries it is entailed in the economic power of the EU as a block of twenty-seven integrated economies. Further, EU law as international law itself constitutes a growing body of law. Yet, the doctrine of external autonomy constitutes another powerful instrument for the Court to ‘Europeanise’ international law. Firstly, EU law also often determines whether and to what extent the decisions of international courts and tribunals will be received in twenty-seven states, as well as the rules of engagement for Member States with those bodies. For instance, the decisions of the Court in Achmea, Mox Plant, and Opinion 2/15 all share the characteristic of being a ‘silver bullet’ for, respectively, intra-EU ISDS arbitrations, intra-EU inter-state arbitrations under international agreements, and the inclusion of investor-state dispute settlement mechanisms in EU FTAs. Secondly, by requiring that international courts and tribunals decide the disputes before them in conformity with a growing number of domestic rules and values. For instance, in Opinion 2/13 the Court objects to the Draft Agreement would breach the autonomy of the EU to set up different, higher standards of protection of human rights on its territories. Similarly, in another case not discussed in this Ph.D. of Kadi, the Court substantially created a justification for the incompliance with a resolution of the Security Council based on Chapter VIII of the UN Charter on the basis of ‘domestic’ standards of protection of fundamental rights. It is also striking how in most of the decisions examined the Court seems to base its assessment of the compatibility of international courts and tribunals with EU law on the basis of the values underlying the new primary law right of judicial protection as today is set out in article 19(1) TEU and 47 CFREU. It was seen in Chapter […] above that such right, as it has been interpreted and applied by the CJEU after Lisbon, is no longer regarded as a minimum standard of enforceability of EU law necessary for the EU to operate autonomously. Rather, it has been interpreted by the Court as to encompass EU-specific substantive standards of justice derived from article 2 TEU, which national courts must meet for the sole reason of their participation in the EU judicial system.
3. Issues with the approach of autonomy as primacy
An interpretation by the Court of the doctrine of external autonomy as a tool to affirm primacy of the EU over international courts and tribunals is questionable for two main reasons: firstly, due the different circumstances the characterise the Court’s relationship with national and international courts and tribunals. Secondly, the impact on the internal constitutional structure, legitimacy, and overall role of the EU in the external relations that arise from the decisions of the Court.
3.1 The different considerations that underlie the relationship of the Court with national and international courts and tribunals
Starting with the first point, an approach of the Court to autonomy as ‘primacy’ over international law disregards the fundamentally different considerations that underlie the internal and external relations of the EU. The relationship between the EU and its Member States is based on a voluntary commitment to participating in a common project for the benefit of the “peoples of Europe,” which, in the areas of competence of the EU, authorises the EU to effectively ‘merge’ the legal orders of the Member States into a separate legal system, with different characteristics, different interests, and different aims. It is also on the basis of the consent of the Member States that the Court can play its ‘supraordinate’ role as a Court of Appeal, Court of Cassation, and Constitutional Court of the EU in relation to their domestic courts. On the contrary, international law is made of a plurality of equally ordinated actors, which all advance their own agenda and interests, and possess their own characteristics. The courts and tribunals operating in international law are set up in the framework of different legal regimes, don’t share a common system of values and principles, nor they operate on the basis of the common political willingness to create a common identity for their users, as it is the case in the EU. They often are even restrained by express rules – such as the prohibition to create precedent, or the creation of specific standards of the rule of law – from doing so. As the Court itself acknowledged in Opinion 1/17, relationships among members of the international law community, courts included, are thus based on a delicate balance of reciprocal concessions, in which the Court should meddle only when strictly necessary to protect the essential traits of the legal order from immediate and irreparable consequences. On the opposite, an approach of primacy of EU law over the operation of international courts and tribunals is liable to create wrong analogies with the internal relations, transforming EU law in the default standard of rule of law for international courts and tribunals and the position of the Court as a mover of the development of the international relations of the EU. For that reason, external autonomy can never be translated into the creation on the part of the Court of a hierarchy of sources and relationships in which the EU treaties sit on top. It should more accurately be limited to defining a specific ‘way of being’ of the EU according to certain fundamental values and external objectives pursued in the external relations, under the vigilant eye of the CJEU, which international courts and tribunals need to consider when exercising their outer scrutiny.
3.2 Negative impact on the internal constitutional structure, legitimacy, and overall role of the EU in the international relations
An approach of the Court to autonomy as primacy also negatively impacts on the internal constitutional structure, legitimacy, and overall role of the EU in the international relations, ultimately undermining its capacity to create the “peoples of Europe.” It was seen above that the approach to outer scrutiny in the EU treaties and in the EU external policy makes of openness towards international dispute settlement mechanisms an element of the identity itself of the Union, which binds all EU Institutions, Court included, to act in a certain manner. While deviations from the approach required in the EU treaties may not be immediately enforceable – due to the breadth of the guidelines set out therein – unjustified and persistent divergences may undermine those acts adopted in derogation from them. Broadening a question formulated by Sarmiento a few years ago in relation to fundamental rights, the issue that appears to arise from an approach of the Court to external autonomy as primacy is whether “[B]y failing to deliver the goods in such a strategic but also “existential” area […], is the Union risking total failure?” While this Ph.D. agrees with the conclusion also reached by the author that “the big malaise […] does not justify an overreaction questioning the Union itself [as] never has the Union been more necessary than today, and it will continue to be crucial for our existence as Europeans,” one may argue that this conclusion can only represent an interim response valid for a limited period of time. The prolonged failure of the Court to deliver on the implementation of the identity of the EU in the external relations as a supporter and promoter of the rule of law through outer scrutiny of international courts and tribunals has deep systemic repercussions of its decisions and may indeed eventually damage the project of “ever closer union.” This conclusion seems to be supported by five main arguments.
a. Issues with the credibility of the EU in international law
To begin with, a default approach of ‘prevalence’ of EU law over international commitments of the EU and its Member States relating to international courts and tribunals undermines the international credibility of the Union as a reliable international partner and, consequently, its capacity to use international law as a tool to both safeguard its interests and values internally and advance them externally. An example is Opinion 2/15 discussed above, on the allocation of competences in the FTA between the EU and Singapore. On that occasion, adopting a ‘negative’ understanding of autonomy, the Court disregarded its previous case law and concluded that the inclusion of ISDS in the trade agreement triggers an obligation for the EU to conclude the EUSFTA as a mixed agreement. The outcome of that opinion – that agreements including an ISDS mechanism must mandatorily be ratified by both the EU and its Member States – ultimately impacts on th e external policy of the EU by significantly reducing the appetite for including such mechanisms in its future international treaties. Despite the subsequent revirement of the CJEU, the impact of the Court’s decision has however already had important policy repercussions on the credibility of the EU in the eyes of its negotiating partners: it substantially forced the reopening of negotiations lasting over 10 years, for which the Council had already given a mandate, and after a final agreement on the text had been reached. Further, faced with the perspective of FTAs having to undergo long and unpredictable ratification proceedings before national parliaments for the whole of the EU FTAs, the Commission formally ‘split’ trade and investment agreements. As an effect, the EUSFTA is now divided in an FTA and an Investment Protection Agreement (IPA). Future FTAs and IPAs may also be negotiated in parallel, but separately. Arguably, in so far as it reduces the scope of the EU agreements, this escamotage of this Commission to reduce the negative effects of the Court’s decision over the negotiation of EU FTAs weakens the negotiating position of the EU, reducing its bargaining power vis-à-vis its trade partners. This became visible in the negotiations of the EU-Japan Economic Partnership Agreement (EPA), which was being negotiated at the time of the decision: after, as an effect of the Opinion 2/15, the contracting parties eliminated the ISDS mechanism included in the EPA and currently, Japan has today lost any incentives to open separate negotiations in investment protection, given its very different position on ISDS reform against the ICS and the multilateral Court that the EU was previously negotiating.
b. Issues with the legitimacy of the EU vis-à-vis national and international courts
Secondly, an approach of autonomy as primacy is liable to undermine the legitimacy of the EU vis-à-vis national and international courts. The reactions to Opinion 2/13 by Member States and the ECtHR are exemplificative: as Sarmiento notes, the outright refusal of the Court to be subject to the control of Strasbourg in Opinion 2/13 made frown national constitutional courts, that have experienced years of subjection to the control of both the CJEU and the ECtHR. Sarmiento makes specific reference to an order adopted by the German Constitutional Court relating to the implementation of the European Arrest Warrant issued by the Italian authorities. Through that order, the German Constitutional Court for the first time reviewed the doctrines of mutual trust and primacy to safeguard “the constitutional identity guaranteed by the Grundgesetz” and declared that the level of protection of human rights of the CFREU may not fall below that safeguarded in the Convention. Notably, the national court rejects the Melloni-line of case law of the CJEU – providing that the level of protection of fundamental rights in cases fully governed by EU law is determined by article 53 CFREU – on which the Court had expressly relied upon in Opinion 2/13 to determine the incompatibility of the Draft Agreement with EU law. Opinion 2/13 also negatively impacted on the relationship between the CJEU and the ECtHR. In what seemed to be a clear ‘warning’ to the CJEU, the President of the ECtHR declared in a speech given a few weeks after the publication of that opinion that “the important thing is to ensure that there is no legal vacuum in human rights protection on the Convention’s territory, whether the violation can be imputed to a State or to a supranational institution.” These words make clear that, regardless of the Union’s accession to the ECHR, the case law of the CJEU on the protection of fundamental rights in the EU would not be exempted from scrutiny against the rights in the Convention before the ECtHR.
c. Issues with institutional balance
From another standpoint, the interpretation of the Court of autonomy as an affirmation of primacy often appears to overlook the institutional balance set out in the EU treaties. This happens when, by curtailing the space for voluntary acceptance of some cession of sovereign powers to international dispute settlement mechanisms of the Commission and the Council, the CJEU heavily limits the necessary space for compromise that they need to develop the international relations of the EU. As discussed, participation in international courts and tribunals requires some degree of cession of sovereign rights, achieved through compromise on the operation of certain rules of the domestic legal system. Within the framework of the EU treaties, the decision to voluntarily accept those rules lays with the political and executive Institutions of the EU. The role of the Court is only to patrol that that cession does not result in serious violations of the constitutional framework. Autonomy as primacy instead legitimizes a pervasive scrutiny on the part of the Court over the exercise of powers of both the Institutions and the Member States in the external relations. While one may agree with Cremona that the Court normally tends to safeguard the policymaking space of the EU, in particular vis-à-vis the Member States, the case law above evidences that the Court at times meddles with highly political issues, which normally belong to the policy-making institutions of the EU. That became evident in the discussed case of Opinion 2/13. On that occasion, by trumping the EU accession to the ECHR, the Court effectively intruded in a political decision that the Member States had not only agreed upon, but also expressly “constitutionalised” in article 6(2) TEU. It was evidenced above how that decision of the Court entailed a high degree of scrutiny over the arrangements set out in the Draft Agreement to question whether any meaningful outer scrutiny on EU law may ever be compatible with EU law. In so doing, it “seems to be paying little attention to inter-institutional balance when deciding on the scrutiny of its review.” So that one may not but agree with Lazowski and Wessel, who argue that it is unclear “to what extent the principle of autonomy can be squared with a voluntary acceptance of external norms.” As Douglas-Scott also notes, the practical outcome of that approach is that, while the Commission is open to an infringement action if it fails to complete the process of accession, the Court arguably irreparably compromised the fragile political compromise withstanding it: if accession remains a possibility, the Court’s opinion makes it in substance very unlikely that a sufficient consensus will be reached among the EU and the Council of Europe’s members to trigger a second procedure. Decisions such as Opinion 2/13 thus raise the “question of who the Herren der Verträge (Masters of the Treaties) really are,” and may ultimately endanger the correct development of the constitutional discourse in the EU, as well as damage its international reputation and position envisaged by the EU treaties as a ‘standard setter’ in international law.
d. Issues with the clarity of judicial reasoning
The approach of the CJEU to autonomy as primacy in the external relations further negatively impacts on the clarity of the judicial reasoning and, as a consequence, on the overall legitimacy of the Court. Admittedly, unclear judicial reasoning is oftentimes the inevitable result of the peculiar composition and functioning mechanisms of the Court, which, among other things, do not include the instrument of dissenting opinions. However, it is argued that this situation is worsened by an understanding of autonomy as primacy. Firstly, in these cases the Court neglects its dual nature as both a ‘constitutional’ and ‘international’ court and acts purely as a ‘constitutional’ court when deciding on cases with a marked international law profile. Secondly, in the framework of a constitutional and policy favour for engagement with international courts and tribunals, the Court is pushes the boundaries of the possible interpretations of the law and often takes decisions that contradict its own case law. Two particularly representative cases are Achmea and Eco Swiss. In order to safeguard its perceived importance of competition law in the EU legal system, the Court in Eco Swiss introduced the new concept of ‘EU public policy’ where it could have simply relied upon primacy. That concept was also too broadly defined to include the all-encompassing notion of ‘competition rules.’ The effect of Eco Swiss was to raise questions on both the exact notion of EU public policy and of the scope of the competition law public policy exception to the recognition of commercial awards. Those questions arguably remain, to date, unanswered. In its subsequent decision in Gazprom, the Court indeed didn’t clarify the scope of the competition law public policy exception despite being invited by AG Wathelet, who discussed the point at lengths in his opinion, to do so. As for Achmea, the decision raises more questions than it answers: the decision is indeed firstly unclear in its scope, i.e., whether it applied to (a) only article 8 of the Dutch-Slovak BIT; (b) investment arbitration in all intra-EU BITs; (c) investment arbitration in all intra-EU BITs and investment arbitration under the Energy Charter Treaty; (d) investment arbitration also in the BITs in force between the Member States and third countries; (e) investment arbitration tout-court. That is particularly problematic considering not only the importance of the outcome judgement, but also the face that it was rendered by the Grand Chamber, whose judgment, are “supposed to be the crème de la crème of the case law at the Court of Justice [and] meant to serve as precedents and to provide ultimate guidance to national courts under the preliminary ruling procedure.” Due to the uncertainty raised by the decision, in the wake of Achmea, Member States themselves could also not agree on the scope of that decision when adopting the s.c. political declarations “on the legal consequences of the Achmea judgment and on investment protection,” aimed at setting out commitments for the future treatment of investments in the EU. Nor the discussed clarifications provided by the President of the CJEU during a subsequent conference make up for this lack of clarity. As it also was noted by another speaker later during the same event, that late ‘extra-judicial’ addition cannot, and should not, impact on the understanding of a judgement delivered in a court of law. The effects of the uncertainty created by the Court’s decision in Achmea are widespread and manifest in the numerous actions of Member States – within and outside of the EU – for the set-aside investment of arbitration awards rendered under the umbrella of intra-EU BIT proceedings as well as in the withdrawal of claimants from such proceedings already commenced. The discussed issues with the reasoning of the CJEU may raise serious issues of legitimacy of the Court in the long term: as Lazowski also notes, while the particular functioning of the CJEU has resulted, from time to time, in vagueness, “if relied upon too frequently, it may undermine the superior position held by the Court of Justice.”
e. Issues with the protection of individuals through access to a remedy
Finally, an approach of the Court to external autonomy as primacy questions the key identity trait of the EU as an entity whose development revolved around the protection of individuals. By closing itself to a dialogue with outer courts pursuing similar objectives of protection of individual rights, or even hindering their operation, the Court rejects virtually any outer scrutiny on the effective functioning of the system of judicial protection of the EU. Yet, the resulting ‘competition’ and dialogue among courts – each willing to have their views heard and gain legitimacy with their users – would arguably translate in an enhanced level of protection of individual rights and effectiveness of all dispute settlement mechanisms involved. That approach of primacy towards international dispute resolution mechanism also precludes to the Court the possibility to fully participate in an inter-judicial dialogue, depriving it of the opportunity to enhance the protection of individual rights externally through dialogue with other judiciaries. A reading of autonomy as primacy also results in the Court questioning the existence itself of other dispute settlement mechanisms able to effectively uphold individuals’ rights. This consideration is evident in the position of the Court on the protection of fundamental rights and ISDS mechanisms. In both cases, the CJEU rejects integration and seems to uncompromisingly envisage a binary option for individuals between either the system of judicial protection set out in the EU treaties or no protection all. In both cases, the safeguard of the effective judicial protection of the EU-specific values is considered the preferable option. Firstly, by fencing off intra-EU actions from the scrutiny of the Strasbourg Court and essentially conditioning the accession of the EU to the ECHR to a reduction of the scrutiny that the ECtHR currently already carries out of assessing the breach of fundamental rights among Member States, the Court arguably prioritises the principle of mutual trust over the effective protection of fundamental rights of the individuals, without there being any indication of that hierarchy of values in the EU treaties. Further, the watertight separation between international law and EU law that derives from the approach of the Court of ‘protection’ from international courts and tribunals to the external autonomy of the EU also sits uncomfortably with the role of ISDS tribunals of supporting the achievement of values and rights that the EU champions, including the strengthening of the rule of law, and in pursuing them outside of its borders through its decisions. The point is discussed in more detail below. Overall, an approach of the Court of primacy of EU law over outer scrutiny of international courts and tribunals seems problematic not only because it assumes the self-sufficiency and full effectiveness of the system of judicial protection of the EU – debatable in the light of the recent issues with the rule of law inside the EU demonstrate. It also entails considerations in terms of an outdated ‘superiority’ of the protection of individuals’ rights through the system of judicial protection of the EU that sits uncomfortably with the important role of effective protection that other international dispute resolution mechanisms play today in international law.
4. the issue with primacy of the EU and ISDS
The debatable approach of primacy of the Court to ISDS deserves specific consideration for two main reasons. Firstly, ISDS arbitration – particularly when administered under the ICSID convention – represents the highest degree of ‘delegation’ by sovereign entities to an international dispute settlement mechanism, which only finds a comparable example in the CJEU. ISDS arbitration thus is the embodiment of the dilemma that the Court faces between embracing a dispute settlement mechanisms that pursues the policy and constitutional aim of the EU effectively implementing the rule of law internationally, and accepting limitations to its system of judicial protection. Secondly, the case law of the Court on ISDS mechanisms is peculiar compared to most of the other cases discussed above in so that it its effects support the policy approach of the Council and the Commission to international investment law. Generally speaking, the same conclusions reached until here in relation to the undesirability of the Court’s approach to autonomy as primacy over international law also apply to the case law on ISDS arbitration. However, the following further points deserve attention. 4.1 ISDS and institutional balance
The Court’s general approach to investment arbitration is broadly line with the objectives of the Commission and the Council, i.e., overcoming the ISDS system by way of, internally, terminating of intra-EU BITs (Achmea); and, externally, creation of a different, court-like, system of judicial protection, the ICS (Opinion 1/17). While this work has argued at lengths for a less activist role of the Court in the external relations, and a backseat approach of the Court in relation to policy approaches of the EU is thus welcome – the position of the Court to ISDS remains criticisable from a standpoint of institutional balance for two main reasons. Firstly, while the Court’s approach to intra-EU BITs in Achmea supports the longstanding adverse position adopted by the Commission towards these agreements, that seems to be a mere coincidence rather than an expression of the Court upholding that policy position. This seems to be demonstrated by the numerous references in Achmea to Opinion 2/13 (7 in a 62 paragraph-long decision), which suggest that the CJEU was seeking continuity with its previous case law on autonomy rather than supporting the position of the EU executive. The displease of the Court with ISDS seems indeed rather based on its traditional concern for the safeguard of the integrity of the system of judicial protection of the EU, as it seems to be confirmed by not only Achmea but also Opinion 2/15. This latter opinion, it was seen, revolved around the risk that ISDS tribunals may remove “disputes from the jurisdiction of the courts of the Member States” by giving investors a choice on the avenue through which they can pursue their claims. Secondly, even if the approach of the Court in its decision in Achmea had been connected to the judiciary opting to pay lip service to the executive in observance of the institutional balance set out in the EU treaties, that would not exempt it from deciding the cases before it within the constitutional framework of the EU treaties and carefully articulating its reasoning. To a different conclusion seems to lead the analysis of the decision of the Court in Opinion 1/17, which, as already observed in the previous sections, signals a change of stance of the Court on the compatibility of an external dispute settlement with EU law compared to its previous line of case law. While Court does not expressly explain on what grounds it decided to adopt a ‘lighter touch’ to the assessment of autonomy in relation to the ICS, the reference in that opinion to the principle of reciprocity supports the argument that that decision was meant to support the EU executive in advancing its policy agenda in the external relations.
4.2 ISDS and Europeanisation
This Ph.D. already expressed support for the renewed approach to autonomy in Opinion 1/17. However, building on the arguments set out above, it is interesting that that change of heart of the Court ultimately supports a comprehensive project of ‘Europeanisation’ of international law put forward by the Commission through the imposition of the very European perspective of the ‘superiority’ of courts over tribunals and superiority of the quality of the elements of the primary law right to judicial protection set out in articles 19(1) TEU and 47 CFREU.
a. Courts over arbitral tribunals
As Tietje notes, arbitration has been recognized as an exercise of individual freedom as a human right, and an effective tool of dispute settlement even in disputes that involve the public interest. Therefore, states as much as the EU cannot impose dispute settlement through national courts. Yet, the impression that one gains from reading the relevant cases of Achmea, Opinion 2/15, and Opinion 1/17 is that international investment arbitration is the ‘poor’ brother of a court, which must be ‘redeemed’ by dressing it as such, or better – by eliminating it tout court. That view reflects the very continental approach of favour for courts over arbitration tribunals, mirrored in the Commission’s reform of ISDS arbitration. The Commission has indeed made no mystery of the fact that it considers that the attribution of court-like features to ISDS arbitration in the ICS and in the MIC intends to embrace a “modern and reformed approach to investment dispute resolution” which ultimately will make the system “fairer.” The reference to fairness hints immediately to value-based considerations, particularly the ones included in the primary law right to judicial protection set out in articles 19(1) TEU and 47 CFREU as interpreted and applied by the CJEU post-Lisbon in the internal relations of the EU. That approach emerges, for instance, in the different approach of the Court to the guarantees of independence of the appointments of the decision makers in the ICS and Achmea: the former, declared as undoubtedly compatible with article 47 CFREU, the latter implicitly questioned in the exclusion that ISDS tribunals may be qualified as “courts or tribunals” of the Member States. The Court’s reasoning on the rationale of why the members of the ICS would be more independent than arbitrators seems however unconvincing and ultimately merely linked to an overall bias for court-like dispute settlement mechanisms, such as their appointment by Canada and the EU; the fixed term appointment, compliance with the IBA Guidelines, and the requirement that the ICS’ members “shall not take instructions from any organisation, or government with regard to matters related to the dispute.” Yet, the Court fails to observe that there is a good argument that the members of the court-like ICS are likely to be no more, if not less, independent of the parties compared to the party appointed arbitrators which sit on ISDS tribunals. To begin with, this ICS mechanism of appointment of the members of the ICS creates the problematic situation that only one party to the dispute – the EU or Canada – may have a saying on the appointment of the individuals that will decide it. Nor it can be taken as a given that, in making the appointments, the home state of the investor will necessarily make their interest, as the interest of the investor and the state may be potentially conflicting: proof is that ISDS mechanisms were set up precisely to disentangle the position of investors from the protection of their home states, considered insufficient. When one takes a step away from the very ‘European’ bias towards courts, party-appointment can be really regarded as a ‘super-guarantee’ of independence of the tribunals compared to domestic courts, which arguably makes ISDS arbitration the only one dispute settlement mechanism able to create a level playing-field between state and non-state parties: for the purpose of ISDS litigation, states are not sovereign actors, but rather parties to a dispute as much as individuals. They have the same rights as individuals and can influence litigation (almost) in the same way as individuals do. From a rule of law standpoint, this seems an advantage rather than a criticism of ISDS arbitration, as it allows individuals to hold directly accountable states for their violation of international rules in a truly neutral forum. Further, apart for occasional distortions that characterise all dispute settlement systems – including the system of judicial protection of the EU – the large independence and delegation that characterises the system provides a strong overall incentive of arbitrators is to act fairly, as from that arguably derives their credibility, the legitimacy of their awards, and, thus, their reappointment. It would also not be in the interest of the investor to select an arbitrator that is evidently biased: such a choice would indeed open the door to a series of procedural difficulties, including the risk of challenges, and the disincentive on the part of the host state to comply with the award. Notably, the ground on which arbitrators are chosen is normally not their links with one or the other party, but rather their views on certain legal interpretations among various possible ones. ISDS arbitration also normally provides for various procedural guarantees for the independence of the arbitrators. These include the mechanism of challenges, codes of conduct, incompatibilities based on the nationality of the arbitrator, and the appointment of a super partes chairman to rebalance the ‘views’ of the party-appointed arbitrators. The legitimacy of ISDS is for these reasons heavily linked to party appointments. The Court’s argument that the fixed term appointment, compliance with the IBA Guidelines, and the requirement that the ICS’ members “shall not take instructions from any organisation, or government with regard to matters related to the dispute” guarantee higher independence of the members of the ICS vis-à-vis the traditional system of appointment of the arbitrators in ISDS arbitration seem equally flawed. To begin with, the ‘permanence’ of the appointment and the existence of the retainer fee don’t necessarily entail independence. Strict rules of independence and conflict of interest and the administration of the appointments by institutional bodies independent by states can achieve comparable results, as it is demonstrated by the fact that highly independent dispute settlement mechanisms, such the US-Iran Tribunal – which Alter qualifies as enjoying a large degree of delegation – are not permanent. Indeed, guarantees in terms of nationality and code of conducts for arbitrators already exist virtually under all sets of arbitration rules. Even more, it is almost paradoxical that the text of the CETA makes reference to the one set of guidelines, the IBA Rules, whose widespread acceptance by the international arbitration community almost makes them a legal requirement. Finally, in relation to the CETA requirement that the members of the ICS “shall not take instructions from any organisation, or government with regard to matters related to the dispute” the CJEU also fails to acknowledge that the footnote at the end of the sentence setting up that requirement expressly allows individuals receiving a remuneration from a government to be appointed on the tribunal. The provision was arguably included to allow civil servants to sit as members of the ICS. Interestingly enough, the IBA Guidelines to which CETA itself makes reference, include the situation that an arbitrator or their firm derives “significant financial income” from a party, as it would be the case for a civil servant in relation to the state party, among the “red list” of non-waivable “circumstances that necessarily raise justifiable doubts as to the arbitrator’s impartiality or independence.” In conclusion, by overcoming party-appointment in favour of a list mechanism for the selection of the members of the ICS administered by the contracting parties of the agreement, CETA takes a notable step back from a principled advancement in international litigation and paves the way of a ‘repossession’ by state parties of investment disputes. That conclusion also seems confirmed by other parts of the text of the CETA, including the provision on the limitation of the use of precedent, that will be discussed below, and the staggering provision on the power of the Joint Committee to adopt interpretations of the CETA binding on the ICS. Another ‘court-like’ element that the EU is pursuing in the reshaping of investor-state disputes is the creation of an appeal mechanism. Yet, the pursuit of an international court over arbitration seems to ignore that the main reason underlying the non-inclusion of appeals in ISDS is one of certainty and speedy resolution of disputes. Arguably, when it comes to international economic transactions, a definitive answer is better than a right one: this explains the success of arbitration in both the commercial and investment realms inter alia as a tool to ‘get around’ lengthy court proceedings. It also explains why ICSID, the most restrictive ISDS regime in terms of revision of final awards, remains the most popular one, even for the EU. A general appellate mechanism, such as the one included in the ICS, may instead duplications in the lenght, cost, and complexity of the arbitral process as it is further likely to turn appeals into the norm: the losing party would arguably be inclined to file an appeal for mere reasons of accountability vis-à-vis its stakeholders (for companies) or citizens (for states). As noted during the discussions on the inclusion of appeal mechanisms in state-to-state arbitration, the precursor of ISDS arbitration, one should thus take appeals in international disputes with the outmost care, as they are capable of hindering not only the effectiveness of the regime of dispute settlement, but its legitimacy, opening the door to abuses for the achievement of political aims. It is also paradoxical that the court-like features of the ICS – and even more so the MIC – may render them more rather than less invasive of the autonomy of the EU compared to ISDS tribunals. This due to the centralization of investment disputes in a two-tier decision-making body, capable of creating authoritative precedent and influence the development of the standards of the international rule of law on which the exercise of sovereign powers – including those of the EU – will be assessed going forward.
It should be noted that, while there undoubtedly is a tendency of arbitral tribunals to cross-refer to each other, investment arbitration is not built around a ‘doctrine of precedent’ or ‘stare decisis’ at least as understood in common law. According to this doctrine, courts must follow and apply the law as set out in previous decisions of higher courts with a view of developing a coherent body of decisions that develops the system of law. This was connected by some commentators to article 53 of the ICSID Convention, providing that awards “shall be binding on the parties.” In reality, a doctrine of ‘precedent’ appears incompatible with the nature itself of international investment law and arbitration, as a fragmented regime made of a plurality of legal frameworks, and each including different standards of the rule of law enforced by ad hoc tribunals set up on a case by case basis. International investment law, and arguably international law in general, lack the hierarchical element that characterizes the common law-based stare decisis doctrine. They also arguably don’t operate in a ‘system’ of law characterized by a homogenous set of principles and values in the framework of which the case law has a function of development: investment treaties are set up to pursue specific aims (i.e., incentivize investments), and arguably the parties have not intention to for those aims to guide the development of the law beyond what is necessary for their effective implementation. Different from ‘precedents’ are the references to ‘previous consistent decisions,’ which arguably are a necessity for international dispute settlement mechanisms. To begin with, those mechanisms operate under bodies of rules that only provide the dispute resolution body with the broad lines of the legal framework under which adopt their decisions. Secondly, they make use of the same rules of customary international law to do so. Reliance on previous consistent decisions in international dispute settlement bodies is also supported by a general quest for legitimacy of both dispute settlement mechanisms and the parties themselves, as a tool to support the meaning that they suggest conferring to the broad provisions that they are called to interpret. Article 38 of the Statute of the ICJ indeed identifies “judicial decisions,” along with scholarly writings, as “subsidiary means for the determination of rules of law.” Reference to previous consistent decisions has finally been regarded as a necessity, or even a ‘duty’ of international decision makers in a key to contribute to the harmonious development of investment law and increase certainty of the rule of law. Indeed, because of their connection with the rule of law, coherence and predictability are one of the concerns identified and currently being addressed in the context of UNCITRAL Group III on ISDS reform, with the ICS and MIC being identified as one of the possible solutions. From this standpoint, one may even argue that reference to previous consistent decisions is an essential element of the necessary ‘delegation’ referred to by Alter for independent decision making bodies to effectively exercise outer scrutiny on the actors in international law by ‘autonomising’ them from their creators and allows for the creation of a direct link with those on which the decisions will impact. It is perhaps indicative that the Statute of the ICJ, the ‘first generation’ court par excellance expressly excludes that its decisions may have “no binding force except between the parties and in respect of that particular case.” The creation of the ICS and the MIC is likely to enhance the mentioned tendency of resorting to previous consistent decisions in investment disputes, making of these two bodies powerful interpreters and enforcers of international standards of rule of law in the field of investment law, able to create competing standards of the rule of law on the basis of which sovereign powers, including the EU’s, will be assessed going forward. This is due, firstly, to the ‘legitimacy’ that the ICS and the MIC will have from the outset. The former is part of a new generation of investment agreements encompassing state of the art standards of the rule of law in international law, while the latter will be a ‘world court’ for investments. Secondly, the higher volume of cases decided by one settlement mechanism formed characterised by a two-tier jurisdiction is likely to contribute to its legitimacy and visibility vis-à-vis national and international courts. Admittedly the EU and Canada have set up a number of provisions to seek to limit this phenomenon, in the context of the discussed effort ‘repossess’ international investment law and arbitration. Notably, awards issued by the ICS (and the same will probably apply to the MIC) are said to be binding only between the disputing parties and in respect of that particular case. CETA parties’ also have sought to safeguard their ‘right to regulate’ against potential investment claims; reserved for the CETA Joint Committee a power to issue binding interpretations of the treaty; carefully spelled out the content of the investment standards set out in the CETA, including the ‘vexed’ standard on non-discrimination, expropriation, and FET; and made sure to limit the effects of the ICS (and MIC) to pecuniary damages (and interest), except, under certain conditions, for restitution of property. Yet, it remains to be seen whether and to what extent these efforts will be successful. Notably, a similar provision in the ICJ Statute that its decisions shall have “no binding force except between the parties and in respect of that particular case,” has not impeded the World court to heavily influence the development of international law through references to both its own case law, and the case law of other international courts and tribunals. From another standpoint, the efforts of the parties to the NAFTA to restrict tribunals’ interpretation of the FET clause have been utterly unsuccessful.
b. The system of judicial protection of the EU over the international protection of individuals right
Section […] above argued that, by depriving individuals of a choice to access to other routes of protection of their rights, the Court disregards the identity trait of the EU as an entity centred on the protection of individuals and assumes the superiority of the system of judicial protection of the EU over other international dispute settlement mechanisms. To the considerations expressed above, it will be added here that this is particularly the case for ISDS tribunals. As mentioned, ISDS arbitration is one of the most advanced expressions of the post-Westfalian movement towards a more individual-centric conception of international law, which champion the rule of law and limit the ‘rule of power’ of states through a system of norms and “remind us that the sovereignty of the State is not boundless.” The ‘protective’ notion of autonomy of the Court in relation to ISDS arbitration overlooks that, oftentimes, this instrument is essential for advancing individuals’ rights to judicial protection, rather than curtailing them. Firstly, ISDS arbitration is a unique international dispute settlement mechanism in so that it allows non-state actors access to an independent mechanism of protection on the same footing as states for the protection of rights universally recognized as part of the rule of law, such as effective judicial protection, non-discrimination, and right to private life and property. Further, ISDS arbitration also arguably entails enhanced protection for EU investors abroad, even beyond the reach of the jurisdiction of the CJEU. In addition to this, in cases where national judicial systems fail – and that may well happen also within the EU – the route of investment arbitration also appears to be more advantageous for foreign investors compared to the customary international law system of diplomatic protection, in so that it affords them “direct access to international arbitration, avoid[s] the political uncertainty inherent in the discretionary nature of diplomatic protection and dispense[s] with the conditions for the exercise of diplomatic protection.” As discussed above, in the face of the ‘re-possession’ of states of investment disputes, the ISDS route is probably preferable for investors over the ICS and MIC envisaged by the Commission. In this context, a provision of the CETA that seems particularly problematic for the protection of the individuals’ rights of the investors regards the power of Canada and the EU to interfere with the decision-making activity of the ICS. According to 8.31(3) CETA, the CETA Joint Committee may adopt interpretations of the CETA agreement that are binding on the ICS, effectively allowing the state parties to readjust the legal framework in their favour on the basis of the experience of the litigation that will be brought under CETA. This provision is also questionable in so that – by setting out an option for the CETA Joint Committee to choose a date from when an interpretation shall have binding effect on the ICS – the agreement implicitly recognizes that these interpretations may be made retroactive during the course of a pending litigation, in clear violation of the right to a fair trial of the investor. Coupled with the abolition of party appointment and a potentially high degree of control over reliance on consistent previous decisions, individuals emerge as the great losers the “repossession” effort of the EU of investment arbitration, which risks undoing fifty years of advancement in the field of investment protection. Finally, the protective approach of the Court towards ISDS arbitration also overlooks that this may at times be the only forum for foreign investors to protect their rights under the relevant investment agreement. The conferral of effects to international agreements in national law, which only can give jurisdiction to national courts to safeguard the rights of investors under investment agreements, is indeed a mere choice in the hands of states, which can decide to expressly deprive them of that effect. Further, even when national courts have the jurisdiction to vindicate the rights of international investors under investment agreements other issues, such as the independence and impartiality of those courts, including in the EU, may in practice deprive investors of tools of judicial protection. As Chase observes, investment arbitration clauses – normally framed within the strong enforcement regimes of either the New York and ICSID conventions – thus are a strong guarantee of the rule of law principle that rights expressly conferred upon individuals will be enforceable. For these reasons, an approach of the Court of primacy of EU law over ISDS arbitration arguably brings into question the position itself of the EU over the safeguard of the rights of individuals.
5.conclusions
This chapter argued that the Court adopts a very high threshold of review of the compatibility of outer scrutiny of international courts and tribunals with EU law. That approach sits uncomfortably with the constitutional identity of the EU treaties and the EU external policy creating a breadth of negative consequences for the functioning and legitimacy of the EU legal system. In particular, section […] observed that the use of external autonomy by the Court as a tool to protect the EU constitutional system from interferences of international dispute settlement mechanisms is only admissible in certain exceptional circumstances, as an expression of the sovereign powers of the EU Institutions to safeguard the fundamental tenets of the legal system of the EU. However, the section went on to evidence that this approach ceases to be compatible with the constitutional framework of the EU and its policy stance of openness towards outer scrutiny when it becomes the default rule. Section […] observed that by normally privileging the safeguard EU law over any interferences, even potential and remote, of international courts and tribunals, the Court adopts an approach to autonomy that closely recalls the use of the primacy used in the internal relations in relation to national courts and tribunals. The approach of the Court to autonomy as primacy seems to emerge in two main ways. Firstly, in certain cases, the Court seems to use autonomy as a tool to expand the scope of EU law (and of its own jurisdiction) at the expenses of the jurisdiction and operation of other dispute settlement mechanisms, and thus derive from that the primacy of EU rules over international law ones. Secondly, in the examined cases, the CJEU shows a tendency to use external autonomy to ‘Europeanise’ international law, demanding that other courts and tribunals conform to EU law in their operation and decision-making activity. Among other things, it was seen above the Court seems to unduly measure the compatibility of international courts and tribunals with EU law on the basis of the EU-specific substantive standards of the primary law right to judicial protection enshrined in articles 19(1) TFEU and 47 CFREU. A right which the CJEU developed as a specific expression of primacy and loyal cooperation of national courts in the new value-based framework of the treaties post-Lisbon. Such an approach of primacy, it was maintained in the following section […], appears misplaced for two main reasons. To begin with, it disregards the different considerations underpinning the internal and external relations and, particularly, the existence in the internal relations of consent of the Member States to limit their sovereignty in favour of the creation of a new legal system based on a common ideological blueprint. It is indeed that consent that legitimises the CJEU to act as the Appellate Court, Court of Cassation, and Constitutional Court of the EU. That consent lacks in the international law context, formed of a variety of equally ordinated dispute settlement mechanisms, set up in different legal frameworks, that interact on equal footing, and don’t have jurisdiction to work towards the creation of a common legal system. Secondly, the approach of the Court negatively impacts on the constitutional structure, legitimacy, and overall role of the EU in the international relations. Section […] specifically analysed the issues arising from the approach of primacy of EU law and its system of judicial protection in relation to ISDS arbitration as a further expression of ‘Europeanisation’ of international law. The section firstly observed that, while the effects of the approach of ‘protection’ of the Court towards ISDS arbitration is shared by the EU Commission and the Council, that approach remains nevertheless criticisable in so that it really reflects the Court’s view of autonomy as primacy, remains at the odds with the EU treaties, and was poorly reasoned. Secondly, section […] criticised the Court’s support in Opinion 1/17 for the ICS system as an expression of a more general, unfounded, bias for ‘courts of law’ over ‘arbitral tribunals’ and for unreasonably undermining ISDS arbitration as a mechanism for effective protection of individuals’ rights. The analysis carried out in this chapter was not meant to be a plea for complete openness towards international courts and tribunals, and particularly ISDS arbitration on the part of the Court. As mentioned, it is part of the definition itself of possessing ‘sovereign powers’ that the EU, also through its courts, should be able to limit the impact of foreign bodies and foreign law in case of inadmissible interference with its nature. However, it is a plea for a shift in the understanding of the Court of external autonomy that is compatible with the constitutional framework of the EU treaties and the policy approach of the EU as a whole towards outer scrutiny. If a ‘negative’ understanding of autonomy was to an extent understandable and necessary at the beginning of the process of formation of the EU – which, as a “strange animal,” needed to differentiate itself from national and international law – this is no longer the case for a mature Union, confident about its role and characteristics. Only by taking stock of this evolution and moving away from autonomy as ‘primacy’ will it be possible for the EU to play the role that the EU treaties have assigned to it: that of a global standard setter, a leader in the advancement of the rule of law, and, most importantly, an entity able to fully live up to its ideological blueprint of creating the “ever closer Union” for the benefit of the peoples of Europe. The next chapter will attempt to draw a new framework for an understanding of autonomy on the part of the Court that is more compatible with the EU treaties and provide some practical suggestions on how to implement it.
CHAPTER VII – TOWARDS CONTROLLED OPENNESS
1. introduction
The previous chapters argued that an analysis of the case law of the CJEU on the compatibility with EU law of international courts and tribunals and their decisions evidences that the Court seems to view ‘autonomy’ as a tool to affirm the primacy of EU law over outer scrutiny of international dispute settlement mechanism. Such approach, it was argued, is at the odds with the important role that the outer scrutiny of international dispute settlement possesses both in the EU treaties and as an essential element of the policy approach of the EU towards international law. It thus creates a number of issues of legitimacy and effectiveness of the EU in external relations. This chapter finally seeks to bridge the gap between the constitutional role that international law and outer scrutiny of international courts and tribunals possess in the EU treaties and in the EU external policy and the current approach of the Court to them. It does so by defining a new paradigm of external autonomy, defined here as ‘controlled openness,’ that is more compatible with the EU treaties. This chapter then also puts forward a series of practical suggestions on how to implement it. It is argued that autonomy as ‘controlled openness’ would not only be more compatible with the EU treaties and address the legitimacy issues previously observed. It would also unleash the full potential of external autonomy as a tool of cohesion and advancement of the EU legal order by supporting the full achievement of the ideological blueprint of the EU of an “ever closer union among the peoples of Europe.” The chapter is organised as follows. Section […] firstly sets the broad lines of the meaning and the advantages of the proposed new paradigm of autonomy as ‘controlled openness.’ Sections […] to […] then put forward some practical suggestions as to how the Court should proceed to achieve it. These suggestions are logically organised from the stage of identifying potential incoherence between EU law and outer scrutiny of international courts and tribunals (3), to handling such incoherence (4), to the tailoring of the final outcomes of the Court’s decisions on the constitutional interactions with international dispute settlement mechanisms (5). Inspiration for those standards was drawn from the vast body of literature on defining and overcoming fragmentation in international law and EU law. Some thoughts on the approach that the Court should take in relation to that dispute settlement mechanism are included in section […]. Finally, some brief conclusions follow in section […]. Before proceeding further, a note of caution is due. The proposals that follow are only indicative and do not attempt to cover the universe of possible choices that the CJEU faces when undertaking the complex task of deciding on the constitutional interactions between the EU and outer scrutiny of international dispute settlement mechanism.
2. autonomy as ‘controlled openness’
A shift in the current understanding of autonomy on the part of the CJEU firstly requires the creation of a paradigm in which conflicts between the EU system of judicial protection and international courts and tribunals can be assessed and managed in a way that is compatible with the EU treaties and the external policy of the EU. That paradigm, is defined here as ‘controlled openness.’ That definition is regarded as capturing the delicate balance underpinning the development of the constitutional interactions between the EU system of judicial protection and international dispute settlement mechanisms in the EU treaties. On the one hand, it reflects the need for a shift in favour of enhanced ‘openness’ towards outer scrutiny and acceptance of the need to cede, for that purpose, some of the sovereign powers of the EU and its Court. The previous chapters evidenced how interactions with international courts and tribunals have an increasingly central role in EU law, turning from an element from which to ‘shield’ the internal market in the early stages of the Community, into a powerful sword to both advance the interests of the EU and live up to the ideological blueprint of the EU treaties of upholding the rule of law internationally. It was also evidenced how arbitral tribunals play a particularly important role in the renewed constitutional framework as one of the preferred tools of the EU to manage its international relations. On the other hand, as Lenaerts notes that approach of ‘openness’ needs to be ‘controlled’ to preserve the constitutional identity and functioning of the EU. The Court has affirmed on multiple occasions in its case law on external autonomy that the constitutional nature of the EU constrains the legitimacy of its action within certain ‘boundaries.’ For instance, it would arguably be hardly acceptable for the Member States and the “peoples of Europe” if today the Union were, for instance, to enter into agreements condoning or, worse, violating, fundamental rights such as access to justice or fair trial. That consideration is particularly important when it comes to the preservation of the EU system of judicial protection, which, as the Court has repeated numerous times, is a cornerstone of the constitutional identity of the EU. Autonomy as ‘controlled openness’ thus does not in any way intend to recommend a complete surrender on the part of the Court of either its control over the Union’s interactions with international law or its relationship with national courts. Rather, the key that characterises the proposed paradigm of autonomy as ‘controlled openness’ – and differentiates it from the ‘old’ approach as an affirmation of primacy – is in the balance that the Court is required to draw between the two elements of ‘openness’ and ‘control,’ which, as suggested in chapter […], should be set in favour of a presumption of openness and dialogue with international law and international courts and tribunals carrying out outer scrutiny over the EU legal system. It is thus argued that autonomy as ‘controlled openness’ supports a redefinition of the concept of external autonomy of the EU legal system as the capacity of the block to define its internal and external action in accordance with its own values and constitutional structure, yet rejecting absolute self-determinism in favour of dialogue, cooperation, and engagement with international dispute mechanisms carrying out outer scrutiny of the EU. A paradigmatic shift of autonomy towards an approach of ‘controlled openness’ is not just a theoretical exercise but entails a number of practical consequences, discussed below. Each of them brings autonomy more in line with both its philosophical understanding and the Union’s constitutional standards and external policy in relation to dispute settlement mechanisms discussed above. Firstly, autonomy as ‘controlled openness’ requires an acknowledgement that the CJEU is but one of numerous courts operating in international law and that coexistence with those courts does not represent a ‘necessary evil’ for the constitutional structure of the EU. Rather, a collaborative interaction with them is an essential instrument of legitimacy of the external action of the Union. That notion is coherent with the original meaning of autonomy developed by Aristotle, which does not merely come down to a complete absence of coercion or external control. Autonomy instead encompasses the idea that an entity may decide to shape and alter its preferences, as well as adopt and implement its norms, in a manner that is independent from others. As discussed in chapter […], the concept of autonomy thus necessarily entails coexistence (autonomous from what?) with others and cannot exist in the absence of that comparator. Along that line, chapter […] evidenced that the dialogue of the CJEU with both national and international courts has contributed to both shaping the identity of the “new legal order” of the European Union and confer legitimacy upon it. Secondly, the new paradigm of autonomy as ‘controlled openness’ means accepting that some degree of interference of foreign courts and tribunals in the EU legal system is not only an inevitable consequence of the plurality of actors in international law. It is itself an expression of autonomy as self-determination of the EU to decide to engage with international law. Theoretical accounts of autonomy support that stance. Dworkin, for instance, separates ‘freedom’ – understood as absolute rejection of any interferences external to the ‘self’ – from ‘autonomy’: if freedom is a negative choice made by an entity at certain points in time to reject external influences, autonomy is a positive way of being, which attains to the choice by an entity to opt whether and to what extent to be influenced in the decision-making process. Consequently, while some degree of independence and self-determination from others is still necessary for an entity to be able to determine its course of action, it does not itself exhaust the concept of autonomy. Rather, in so that it reflects the capacity of that entity to decide whether to find compromise with outer bodies, ‘autonomy’ is evidenced – and not endangered – by the decision to interact with others and subject oneself to certain rules. That approach to autonomy is reflected in the EU treaties and the external policy of the EU, which make of the high degree of engagement with international dispute settlement mechanisms carrying out outer scrutiny over the EU an element of the identity itself of the EU, bound to orient the functioning of the EU, and support the legitimacy of their actions. On the one hand, that high degree of cooperation and dialogue with international dispute settlement mechanisms allows the EU to comply with the rules of acceptance set by the international community for the recognition of its participation in that community. The CJEU appears itself to acknowledge the importance of such recognition and acceptance in Opinion 1/17 on the CETA, where it declares the compatibility of the ICS with EU law “because of the reciprocal nature of international agreements and the need to maintain the powers of the Union in international relations.” On the other hand, engagement with international dispute settlement mechanisms also allows the EU to participate in the definition of those same rules of acceptance for the recognition in the international community and shape them in a way that allows it to advance its values. It was evidenced above how participation in international dispute settlement mechanisms is a key tool for the EU to also safeguard the values, fundamental interests, security, independence, and integrity of the Union even outside of its borders where the jurisdiction of the Court cannot reach, and to advance its values and interests in the wider world.
Finally, the capacity of the EU to set high standards for the protection of rights and values through cooperation and dialogue with other international dispute settlement mechanisms also enhances its reputation in international law: its commitment towards international relations based on the rule of law, and enforceable standards of safeguard of fundamental rights in international law has today conferred the EU a reputation as ‘standard setter’ able to draw other international players towards compliance with its benchmarks of protection. Autonomy as ‘controlled openness’ also requires acknowledging that interaction with international dispute settlement mechanisms may have ‘unifying’ effects on the EU legal system, even when that same interaction fragments the ‘uniformity’ of EU law. In comparative legal theory, interaction with external entities is often seen as contributing to the development and affirmation of their identities by acting as a ‘mirror’ that promotes self-awareness and maximization of rationality through the delimitation of the respective spheres of competence. The idea is that interaction among separate entities allows them, through reciprocal comparison and differentiation, to better understand their own characteristics and values. That also appears to be one of the functions that outer scrutiny of foreign courts and tribunals have in the EU treaties. The analysed case-law of the CJEU on external autonomy arguably evidences that international courts and tribunals functioned in the case law of the Court as an essential tool for the definition of the contours of the autonomous legal order of the EU, i.e., the non-negotiable elements of the Union’s constitutional identity that cannot be compromised on pain of betraying the foundational traits of its legal order. Through conflict and negotiation, that case law delineates the specific traits of a coherent legal system with which individuals and Member States can identify – such as fundamental rights and full judicial protection – and so contributes to the formation of one common identity that alone can achieve the “ever closer Union among the peoples of Europe.” In conclusion, as opposed to a limiting view of autonomy as an affirmation of primacy, the paradigm of autonomy as ‘controlled openness’ embraces an idea of autonomy that views international dispute settlement mechanisms and their outer scrutiny not as a ‘threat’ for the EU, but rather as a tool to safeguard, and even advance, its legal order, and ultimately legitimises its own existence by ensuring that its functioning never betrays certain standards of the international rule of law. External autonomy of the EU as controlled openness thus operates an instrument of growth and definition of the identity of its legal system though mutual exchange with otherness. As mentioned, this of course does not mean that the CJEU should give up all oversight on the growing interactions between the EU legal system and international law or its constitutional relationship with national courts. The constitutional role assigned by the EU treaties to the CJEU of ensuring that those interactions are carried out in a manner that is compatible with their fundamental precepts and don’t endanger the unity of the system remains a crucial feature of the external relations and a guarantee that the ideological blueprint of the “ever closer union” is lived up to in compliance with the fundamental charter of the EU. The ‘openness’ of the EU legal order thus still needs to be ‘controlled’ by the CJEU. However, a more ‘open’ approach of the EU to international law does require a shift in the fundamental understanding of that oversight, by moving towards a presumption of openness towards the engagement with international courts and tribunals carrying out outer scrutiny over the EU. It is argued that such shift, which may be supported by the practical suggestions listed below, would unleash the full potential of the concept of “autonomy” as a tool to ensure greater internal consistency of the EU legal system and reinforcing its role on the external relations.
3. ascertaining conflict
The first step in the implementation of the new understanding of autonomy as ‘controlled openness’ requires a shift in the understanding of the meaning of ‘conflict’ between the EU legal system and international courts and tribunals carrying out outer scrutiny over it. Chapter […] evidenced that the Court adopts an extremely broad understanding of ‘conflict’ which often covers not only actual interferences of international courts and tribunals carrying out outer scrutiny over the EU with the EU system of judicial protection, but also merely potential ones. One of the examples that was given above concerned the Court’s preliminary opinions under article 218(11) TFEU. Chapter […] also highlighted that the Court responds to this broad concept of ‘conflict’ with a principled approach of ‘incompatibility’ between the EU legal system and international law. Evidence of that approach was found in an uncommon ‘high degree’ of scrutiny of the CJEU over the compatibility of international courts and tribunals with EU law and the Court’s seeming tendency to ensure the default ‘prevalence’ of the system of judicial protection of the EU over their outer scrutiny every time that a conflict arises. This chapter argues that autonomy as ‘controlled openness’ requires that only conflicts that constitute real threats to the essential functioning mechanisms of the EU legal system should be considered relevant from a standpoint of external autonomy. That goes through an acknowledgement on the part of the Court that legal issues of an international nature are physiologically governed by a non-hierarchical plurality of sources and normally subject to different regimes and jurisdictions with the possibility of divergent interpretations, without that necessarily raising essential threats for functioning of the EU or its system of judicial protection liable to trigger considerations of external autonomy.
3.1 Normalisation of ‘conflict’
The broad understanding of ‘conflict’ on the part of the CJEU seems to be in line with the ILC Report on Fragmentation, which generally defines it as “a situation where two rules or principles suggest different ways of dealing with a problem.” That notion of conflict encompasses both potential and actual misalignment between concurring interpretations of the same rule and the existence of different sets of rules regulating the same situation. However, the perception of the value of – and approach to – conflict in the ILC report and in case law of the Court are different. In the ILC report conflict is perceived as a physiological consequence of “the spontaneous, decentralized and un-hierarchical nature of international law-making - law-making by custom and by treaty” which traditionally faces courts and operators “with heterogeneous materials at different levels of generality and with different normative force.” The Court seems instead to view any conflict as undesirable and threatening to the external autonomy of the EU legal system. The underlying reasoning seems to be that the concurring interpretations by international courts and tribunal of issues that impinge, even indirectly, into EU matters may endanger the uniformity of the interpretation and application of EU law. This work suggests that the systemic shift in the understanding of external autonomy towards ‘controlled openness’ firstly goes through the ‘normalization’ of conflict, both substantive and jurisdictional, as a physiological expression of the plurality of non-hierarchical sources and regimes of international law. Indeed, not all conflicts between EU and international law, be them of interpretation of the same rule or regarding the actual existence of competing rules and obligations regulating the same situation, threaten the constitutional foundation and functioning of the EU. To that aim, this work supports the adoption of an approach of openness on the part of the CJEU towards the existence of multiple rules – and interpretation thereof – regulating the same situation governed by EU law. Normalisation of conflict responds to an idea of autonomy as ‘controlled openness’ in so that it does not construe the existence of different rules or interpretations or application of the same rules on the part of EU courts and international courts and tribunals in terms of systemic opposition but rather of normal coexistence, each one within its own sphere. That normal coexistence is enabled by the fact that, when dealing with international courts and tribunals carrying out outer scrutiny over the EU the Court should not attempt to construe conflicting interpretations or obligations as part of a hierarchically organised uniform system, as it is used to do in its relationship with the national courts of the Member States. On the opposite, coexistence of different interpretations or obligations governing the same situation across separate legal regimes should be regarded as perfectly normal. An example of how jurisdictional conflicts are normalised in international law is, for instance, the interpretation of rules of domestic law as ‘facts’ on the part of arbitral tribunals. That possibility was discussed, for instance, in the case of Mox Plant, regarding an issue of competing jurisdiction between an ITLOS tribunal and the CJEU. Notably, the ITLOS tribunal formed under UNCLOS to decide the case held that different dispute resolution bodies might differently apply the same rules based on “differences in the respective context, object and purposed, subsequent practice of parties and travaux preparatoires.” The ITLOS tribunal further recognized that the interpretation of norms and legal principles will necessarily depend on the legal framework in which they are applied, so that even identical provisions may be interpreted and applied differently in different legal contexts. “They reflect the differing pursuits and preferences that actors in a pluralistic (global) society have.” An example of how substantive conflicts are normalised in international law can be found in the decision of the investment tribunal in Eastern Sugar, an intra-EU BIT case arisen under the 1991 Czech-Dutch BIT. In that case, the Dutch investor sued the Czech authorities on the grounds that, by enacting a series of three pricing decrees that allegedly discriminated against the investor, the Czech Republic had contravened the standard of fair and equitable treatment set out under the governing investment treaty. The investment tribunal constituted to decide on the dispute rejected the defence put forward by the respondent state that the demands of protection of the investor under international law and EU law were in conflict, or that satisfying the demands of the investor under the former meant disregarding the rules of the latter. Normalising conflict has a number of advantages in terms of shifting towards a paradigm of autonomy as ‘controlled openness.’ Firstly, normalisation of jurisdictional conflict overcomes the objection of the Court that its monopoly of interpretation and application of EU law under article 344 TFEU prohibits in principle other courts and tribunals from dealing with EU law in the context of their adjudicatory activity: if one accepts that multiple interpretations and applications of the same norm across various legal regimes may physiologically co-exist, the fact that arbitral tribunals – and any non-judicial bodies, for that matter – may adopt their own interpretation of EU law does not necessarily threaten the prevalence of the interpretation and application given to those same norms by the Court, even when that interpretation and application has effects within the EU legal system. In so doing, normalisation of conflict simplifies the co-existence of the EU legal system and international law, including, for instance, the much-discussed relationship with the ECtHR and the existence of investor-state arbitration clauses in intra-EU BITs. So, for instance, an approach of normalisation of conflict would render in principle compatible the interpretation on the part of investment tribunals of EU obligations of the Member States for the free circulation of capitals in the context of the assessment of the ‘national treatment’ or ‘most favoured nation’ clauses under bilateral investment treaties. As Davies notes, a shift of perspective towards a redefinition of the exclusive jurisdiction of the Court on the interpretation and application of EU treaties is not constitutionally precluded: article 19 TEU provides general jurisdiction of the Court but does not confer to it a general power of ascertaining its kompetenz-kompetenz. Nor do so articles 267 and 344 TFEU, which merely delineate the powers of the Court. When one accepts the premise that the EU treaties embrace the idea that the EU legal system coexists and operates on different levels with other legal regimes, articles 19 TEU, and 344 and 267 TFEU – often regarded as stumbling blocks to the coexistence of the EU system of judicial protection with other systems of adjudication – only govern internal issues of EU law. They do not extend to other international obligations of the Member States within different legal systems, even though they may also influence their obligations under the EU treaties – be them the ECHR, an investment treaty, or any other international law instrument.
In so that it does not pursue prevalence of one system over another, normalisation of conflict also is a technique of inclusion that avoids systemic a priori clashes, in line with the values and objectives of openness and dialogue with international courts and tribunals carrying out outer scrutiny over the EU envisaged in the EU treaties and in the external policy of the EU. On the one hand, that approach should lead the CJEU to rethink its ‘ownership’ of the EU legal system: while within the EU legal system the interpretations of EU law by the CJEU are still binding on national courts and authorities, normalisation of conflict acknowledges that the EU legal system is also part of international law. Thus, contrarily to the stance taken by the Court in a number of cases analysed above, including Achmea and Opinion 2/13, other dispute settlement bodies outside of that system are entitled to interpret EU law in the context of the jurisdiction attributed to them within their own legal order. On the other hand, normalisation of conflict also accepts that the CJEU may interpret and readapt international law rules to fit the constitutional structure of the EU without prejudicing the interpretation of them that may be offered by other courts and tribunals operating in other legal systems. So, the re-adaptation of customary international law that the CJEU frequently carries out to accommodate the specific characteristics of the EU system of judicial protection, would not cause any issues when appropriately acknowledged and reasoned. Normalisation of conflict also supports the evolution of the constitutional identity of the EU towards unity “of the peoples” behind shared values: in so that it encourages reasoned divergence and dialogue on the non-negotiable elements of the EU legal system, an approach in terms of normalisation of conflict promotes an idea of autonomy as a tool to strengthen a common identity rather than to ensure that the same law is strictly applied within the EU. From this standpoint, in accordance with the EU treaties, external autonomy ceases to be an ‘end’ in itself and instead becomes a mean to fully fulfil the identity of the EU as a tool to achieve the “ever closer union among the peoples of Europe.” Different is the case of conflicts between international courts and tribunals carrying out outer scrutiny over the EU and EU law whereby compliance with their interpretation and application of a certain rule is in its face radically incompatible with EU rules. This issue of conflicts in the ‘narrow’ sense – which are considered in this Ph.D. as the only ones relevant in terms of an approach of autonomy as ‘protection’ of the system of judicial protection of the EU – will be discussed in more details below. Critics may argue that normalisation of conflict will open the door to fragmentation of EU law and to possible incompliance on the part of the Member States with their obligations under EU law. Take, for instance, the varying interpretation of right to access to court or fair trial that may be adopted by the ECtHR in relation to article 6 ECHR, ISDS tribunals in relation to FET clause, or by the CJEU itself in relation to article 47 CFREU. Following this line of reasoning, it may well happen that a domestic court of a Member State deciding on a matter of EU law may resort to the that right as applied by ISDS tribunals or by the ECtHR because, for instance, deemed better suited to that case among a plethora of equally valid possible competing sources or interpretations. Yet, that argument appears unconvincing and unable to offset the multiple advantages of the approach of normalisation of conflict discussed so far. To begin with, it was already observed that normalisation of conflict does not question the internal rules of EU law within the EU legal system, including the monopoly of the interpretation and application of EU law of the CJEU set out in article 344 TFEU, or its internal hierarchy of rules. Normalisation of conflict rather merely accepts that norms and interpretation of norms competing with EU law are equally valid and are not placed in a position of hierarchical subordination to the EU treaties. Therefore, it is perfectly possible that other courts and tribunals outside of the EU may carry out, in the context of their decision-making activity, their interpretation of EU law within their own legal systems. One would certainly hope that those courts and tribunals, as it has often happened in the past, will interpret and apply EU rules as objectively and faithfully as possible and within the confines of their jurisdiction. However, as Eeckhout notes, should they deplorably not do so, it is a fact that their decisions in themselves will not be capable of modifying EU law. In so that it ‘tests’ the loyalty of Member States and their courts, and pushes them and the EU Institutions to face real or merely apparent incompatibilities with rules or obligations arising from outer scrutiny, an approach in terms of normalisation of conflict can even further the project of openness towards engagement with international dispute settlement mechanisms in the EU treaties, as well as the definition of the constitutional identity of the EU. In particular, decentralised interpretation is especially beneficial in so that it supports reasoned exchanges among domestic, EU, and international courts – which are encouraged to provide convincing arguments to support their decisions, for instance analysis in details how the law applies to the circumstances of the case, or the different rationale underlying the specific legal system in which they operate. Those exchanges have the potential to ‘better’ the judicial reasoning of all dispute resolution mechanisms involved, encouraging them to issue more ‘convincing’ interpretations of the law to incentivise the understanding and adoption of their decisions across legal systems. Interestingly, the need for normalisation of conflict has already been acknowledged in the EU legal system. Notably, Advocate General Cosmas in Dior noted that “it is in no way contradictory that the meaning of a provision should vary depending on the matter to which it is applied (in the present case, an intellectual property right), the body interpreting it (in the present case, the Court or the national courts) and the relevant legal framework (in the present case, Community or national).” The CJEU itself seems to have finally accepted the validity of an approach in terms of normalisation of conflict in Opinion 1/17: the Court came to terms with the possibility that the arbitral tribunals in CETA may interpret EU law involved in the cases involving EU law, provided that they do so “as a matter of fact,” i.e., without binding effects on the EU legal system.
The possibility of the coexistence of ‘conflicting’ legal obligations as a result of the existence of a plurality of legal regimes also appears to be acknowledged by the discussed Grandfathering Regulation. Adopting a pragmatic approach to competences, that regulation takes into account both the pre-existing international commitments of the Member States and the political and economic considerations underlying the maintenance of existing BITs or even the negotiation of new ones regardless of the exclusive competence of the EU in the field.
3.2 Conflict in the ‘narrow’ sense as the only one relevant for autonomy
By embracing the idea that the same situation can simultaneously have a variety of interconnected legal implications and interpretations that physiologically co-exist, normalisation of conflict ultimately rejects that all interferences of international dispute settlement mechanisms with EU law may trigger considerations of preserving the external autonomy of the EU legal system. An approach in those terms reduces conflict between international dispute settlement mechanisms and EU law to a handful of very specific instances, where those mechanisms may impose upon the EU or its Member States legal obligations that are irreconcilable with their obligations under EU law. These conflicts, which create a situation whereby “[a]n obligation may be fulfilled only by thereby failing to fulfil another obligation” will be the only ones relevant from a standpoint of their impact on the constitutional interactions of the EU with international law. This type of incompatibilities will be referred herein as conflicts in the ‘narrow’ sense. The case in Kadi, not discussed in detail in this Ph.D., can be regarded as a good example of a conflict in the ‘narrow’ sense: the CJEU was in that case faced, on the one side, with obligations under international law to implement the challenged resolutions of the UN Security Council, and, on the other side, with a treaty duty to safeguard the fundamental rights of the appellants on the EU territories. Both international and EU law appear to support a restrictive interpretation of situations of conflict in the ‘narrow’ sense. On the one hand, the communal nature of international law and the principle of comity among its participants creates a strong presumption against irreconcilable conflict: as Kokasiennemi highlights in the mentioned ILC report on fragmentation, “[t]reaty interpretation is diplomacy, and it is the business of diplomacy to avoid or mitigate conflict.” On the other hand, the approach of openness towards international dispute settlement mechanisms and their outer scrutiny embraced by the EU treaties and the general external policy of the EU, arguably equally supports a presumption against judicial conflict. This restrictive interpretation of conflict in the ‘narrow’ sense could be supported through various tools.
a. Legal duty to attempt a conciliatory reading
Considering that the existence of conflicts in the ‘narrow’ sense is a matter of interpretation, all participants in the system of judicial protection of the EU, the Court in primis, dealing with possible conflicts in the ‘narrow’ sense arguably have a legal duty to attempt a conciliatory reading of seemingly conflicting obligations or interpretations arising from the outer scrutiny of international dispute settlement mechanisms to avoid the situation where compliance with EU law would translate into incompliance with the existence or decisions of those bodies. For instance, by referring to the reciprocal nature of the international agreements of the EU, in Opinion 1/17 the Court concluded that the provisions of the CETA may be interpreted by other courts of non-Member States with which those agreements were concluded or by the international courts or tribunals established by such agreements. Similarly, it was seen above that article 351(2) TFEU codifies an obligation to avoid conflict in the narrow sense, and the CJEU has also previously interpreted the provision in the sense of finding an obligation on the national courts of the Member States to minimize possible incompatibilities between the EU treaties and the international agreements of the Member States “to the extent possible and in compliance with international law, in such a way that it is consistent with Community law.”
b. Actuality of conflict…
A shift towards autonomy as ‘controlled openness’ in line with the EU treaties also arguably requires that conflicts in the ‘narrow’ sense should be limited to cases which present a connection of a certain degree with the present time, i.e., where the international dispute settlement mechanisms is actually creating a conflict in the narrow sense with EU law. Where there is only a risk that a conflict may arise in the future, a decision of the Court in terms of incompatibility of that mechanism or its outer scrutiny with EU law should be a rare exception. On the one hand, it is doubtful that conflicts in the ‘narrow’ sense, which only are relevant for the purpose of autonomy, may arise in cases of merely future and potential incompatibilities with international courts and tribunals or their functioning. Arguably, a strict choice between compliance with competing obligations of EU law or the existence and functioning of international courts and tribunals only exists where two incompatible obligations are applicable at the same time. On the other hand, the CJEU would arguably not be able to carry out the specific analysis of the application and consequences of the competing rules or obligations of EU law and arising from the outer scrutiny of a dispute settlement mechanism necessary to ascertain which one of them should prevail. This arguably requires lowering the degree of interference with EU law that the CJEU accepts, which, it was seen in chapter […], seems currently to be set at the level of ‘mere possibility.’ That ‘preventative’ approach of the Court is typical of the preliminary opinions of the Court under article 218(11) TFEU: for instance, in Opinion 2/13 the accession of the EU to the ECHR was deemed by the Court to be incompatible with EU law on the basis of the fact that “it cannot be ruled out that a request for an advisory opinion [...] might” circumvent the mechanism of preliminary reference, or, with regards to the mechanism of prior involvement of the CJEU in the decision-making activity of the ECtHR, that the agreement didn’t “contain anything to suggest that [the] possibility is excluded” that cases at issue in the proceedings before the ECtHR on which the CJEU has decided should be resolved only by the Court, whose decision should also bind the ECtHR. Similarly, in Achmea the CJEU seemingly suggested that the mere possibility that arbitral tribunals may rule on the interpretation or application of EU law represents a threat to the autonomy of the EU legal order. It may be counterargued that the present proposal of limiting the scrutiny of the Court on the compatibility of envisaged international agreements to actual conflicts would essentially deprive the Court of the constitutional role of preventative control that it possesses under article 218(11) TFEU. The argument could be made that, as these opinions by definition regard preliminary legal texts rather than their actual application, under the proposed new standard of controlled openness a declaration of their incompatibility would essentially only arise in cases of blatant disregard of fundamental rules of EU law, such as an outright ousting of the jurisdiction of the Court in one of its clear areas of jurisdiction. That criticism appears unfounded. While the standard for the declaration of incompatibility of international dispute settlement mechanisms and their outer scrutiny with EU law would admittedly be lowered considerably, the Court would arguably still retain a powerful tool of protection of the legal system that allows it to perform its role under article 218(11) TFEU: that of ‘conditional’ interpretation. Through conditional interpretation, control on the impact of the international dispute settlement mechanisms and their outer scrutiny in proposed international treaties on the part of the Court would not be completely waived, but only deferred to a moment where a conflict actually arises. The section below will discuss the instrument of ‘conditional’ interpretation in more details.
c. …and ‘conditional’ interpretation
Another tool in the hands of the Court to implement autonomy as ‘controlled openness’ regards cases where more than one interpretation of a potentially incompatible norm relating to the potential functioning of international dispute settlement mechanisms is possible. On these instances, the Court has the power to effectively ‘condition’ the compatibility of that norm with EU law to a certain interpretation that avoids fundamental issues of essential rules of primary EU law. Opinion 1/17 provides a good example of how the tool of conditional interpretation can effectively be used by the Court to avoid conflicts in the ‘narrow’ sense. When discussing the compatibility of the proposed draft agreement of the ICS in CETA with the standards of independence of judges set out in the EU treaties, the Court faced the issue that the wording of article 8.31.3 CETA allows in theory the CETA Joint Committee to adopt interpretative decisions of that text with retroactive and direct effect on pending disputes. Such an interpretation of the powers of the CETA Joint Committee would evidently conflict with an essential norm of primary law of the EU, i.e., the right to an effective remedy, in so that it would effectively be liable to retrospectively amend the legislative framework in which those disputes arise, in favour of one or the other party to the dispute. Interestingly, instead of issuing a declaration of incompatibility of the entire ICS with EU law, in Opinion 1/17 the Court effectively conditioned the compatibility of that provision to compliance with EU law. Specifically, the Court held that [T]he consent of the Union to any decision specified in Article 8.31.3 of the CETA will have to comply with EU primary law and, in particular, with the right to an effective remedy enshrined in Article 47 of the Charter. Accordingly, Article 8.31.3 of the CETA cannot be interpreted, having regard to Article 47, as permitting the Union to consent to decisions on interpretation of the CETA Joint Committee that would produce effects on the handling of disputes that have been dealt with or are pending. Privileging this type of conditional interpretation over preventative declarations of incompatibility seems beneficial in two ways. On the one hand, it does not a priori exclude engagement of the EU with an international dispute settlement mechanism on the grounds of merely potential future incompatibilities with EU law. On the other hand, it still preserves the constitutional power of the Court to safeguard the structure and values of the EU from damaging external interferences of international dispute settlement mechanisms. In the case at issue, the conditional interpretation adopted by the Court is a ‘warning’ sign both to the EU and Canada and to the CETA Joint Committee that any applications of article 8.31.3 CETA adopted in violation of the guarantees of effective remedies of the EU may taint the validity of the decisions of the CETA tribunals and raise issues of enforcement before the domestic courts of the Member States. Other cases in which the tool of ‘conditional interpretation’ was used – Opinion 1/92 and the advice of AG Kokott in the framework of Opinion 2/13 – all regard preliminary opinions under article 218(11) TFEU. However, the use of this instrument on the part of the Court in the context or preliminary references or direct actions may also well be hypothesized.
d. From protection of – to trust into – national courts to avoid conflict
In a key of avoiding potential conflicts in the ‘narrow’ sense from a jurisdictional standpoint, it also seems important to review the approach of the Court vis-à-vis the role of domestic courts of the Member States in the EU system of judicial protection. It is argued that, in order to reduce potential conflicts with international law triggered by the safeguard of articles 344 and 267 TFEU, that approach should shift from one of ‘protection’ from external interferences, to one of enhanced openness based on a ‘presumption of loyalty and collaboration’ in the safeguard of the correct functioning of the EU legal system. As Koutrakos notes, the Court has oftentimes identified the existence of conflicts with the autonomy of EU law in the potential impingement in the jurisdiction of national courts by international dispute settlement mechanisms. On those instances, the Court oftentimes issued decisions of incompatibility of those mechanisms with EU law to ‘safeguard’ the position of national courts within the system of judicial protection of the EU provided for under articles 19 TEU and 267 TFEU. That approach of the Court is understandable considering the constitutional importance of the role of national courts as the first points of delivery of EU law and guardians of the effectiveness of EU law: as the CJEU has often repeated, national courts are indeed “indispensable to the preservation of the very nature of the law established by the Treaties.” So, in Opinion 1/09 the Court declared the incompatibility of a unified patent litigation system because, by conferring to the patent court the exclusive right to refer preliminary questions to the CJEU in the field of patents, it deprived national courts to do so and interfered with the mechanism of preliminary reference in article 267 TFEU. Similarly, in Achmea the Court objected to the compatibility with EU law of article 8 of the Dutch-Slovak BIT on the grounds that the finality of investment awards, the freedom of the arbitration tribunal to elect its seat and applicable law, and that national law may limit the extent to which national courts can review such awards. It is also significant that in Opinion 2/15 on the conclusion of the EU-Singapore FTA, the Court justified a departure from its established case law on the ‘ancillary’ nature of dispute settlement mechanism to the substantive provisions which they accompany because investor-state arbitration “removes disputes from the jurisdiction of the courts of the Member States.” While national courts undoubtedly are essential constitutional tiles of the EU system of judicial protection, the position of the Court on the alleged need to ‘safeguard’ their position remains unconvincing from numerous standpoints. To begin with, oftentimes the Court tends to err on the side of precaution and merely presumes that arbitral tribunals or their decision may negatively impact on the jurisdiction of national courts. It is significative that in its decisions the Court rarely analyses the actual implications for the jurisdiction and functioning of national courts of the dispute settlement mechanism whose jurisdiction would allegedly hinder their constitutional role under the EU treaties. So, in Opinion 1/09 the Court omits to refer to the guarantees for the protection of the judicial system of the EU set out in the draft agreement for the envisaged patent system, and, in particular, its limited jurisdiction ratione materiae, which rendered quite unlikely any meaningful negative impact on both the jurisdiction of domestic courts and the EU system of judicial protection. More generally, in that opinion the Court also omitted to mention that the envisaged patent agreement included a number of tools of protection of the judicial system of the EU, as it safeguarded the exclusive jurisdiction of the CJEU on both the interpretation and application of secondary EU law, as well as the interpretation of primary EU law. The patent court system discussed in Opinion 1/09 would have also been bound by the case-law of the Court. Similarly, in Opinion 2/13 the Court overlooks that Protocol No. 8 to the Draft Agreement of the EU to the CJEU expressly preserved the characteristics of the Union and its legal order and provided that accession should not affect powers of the EU Institutions. It also overlooked that article 344 TFEU already confers a monopoly to the CJEU on all disputes between member states regarding EU law and that the ECtHR already has jurisdiction on intra-EU disputes, without that having ever caused any major issues for the legal system of the Union. Further, in Achmea the CJEU did not examine the essential role of support of domestic courts to arbitral tribunals – which may range from the selection of arbitrators to the enforcement of arbitral awards – and the fact that such role would hardly permit in practice that arbitral awards may escape at least some degree of control by national courts that would prevent them from endangering the effectiveness of EU law. Strikingly, Achmea originated itself from a preliminary reference from a national court of the Member States, the German Bundesgerichtshof (Federal Court of Justice). Achmea also evidences the preventative approach of the Court: the CJEU did not engage in a discussion on the different limitations that different existing sets of arbitration rules set to the control of domestic courts over the awards of investment tribunals. The Court thus avoids entering into the discussion – instead entertained by Advocate General Wathelet in his opinion in that case – on how UNCITRAL Rules, the procedural rules applicable in the case at issue don’t impede a review of investment awards by national courts on the ground of public policy. Different considerations apply instead to the ICSID arbitration rules, which instead limit the jurisdiction of domestic court to review arbitration awards from a standpoint of public policy. Along the same line, in Opinion 2/15 the Court did not attempt to justify its departure from its established case law on the ‘ancillary’ nature of dispute resolution mechanisms, which is treated as an assumption. These cases suggest that the assumption of the Court that national courts may need to be ‘shielded’ from external interferences of international courts is hardly compelling. The practice of domestic courts dealing with the existence and decisions of international tribunals adjudicating on EU law matters appears to rather show that not only those courts already possess the necessary tools to stand up, as powerful bastions of protection, against major violations of EU law by arbitral tribunals. They also have proven to be willing to do so. Various decisions of national courts in relation to the enforcement actions brought by the investors in the intra-EU award rendered in Micula support this argument. The case in Micula arose from the adoption by Romania of an array of anti-state aid measures. Those measures were put in place by Romania in view of its accession to the EU on 1 January 2007, which required that the country should adapt its national law to the acquis Communautaire, specifically in the fields of state aid and competition law. When the country withdrew economic incentives for the development of certain disfavoured regions of Romania on the ground of their incompatibility with state aid law, two Swedish investors, the Micula brothers, and their investment vehicles constituted under Romanian law, started an ICSID arbitration against Romania for the violation of various standards of protection entailed in the 2002 Sweden-Romania BIT. The EU Commission participated in the proceedings as amicus curiae in support of Romania. The ICSID tribunal constituted to decide the case eventually awarded the claimants EUR 178 million in compensation for violation of the standards of fair and equitable treatment and minimum standards of treatment. While Romania sought and failed to annul the award before an ad hoc ICSID Committee, the Commission adopted a decision prohibiting Romania to pay the compensation awarded to the Swedish investors, due to that compensation constituting illegal state aid The investors challenged the legality of that decision before the General Court, while at the same time starting a number of enforcement actions under the ICSID convention before EU and foreign courts. The decisions adopted by the national courts of the Member States in the context of those enforcement actions are telling of the tools that those courts possess to safeguard EU law. That is particularly evident from the fact that the award in Micula was rendered in the framework of the ICSID convention, which, it was seen to expressly exclude, per its article 54, that national courts may review awards on the grounds of public policy. On 20 January 2017, Mr Justice Blair of the High Court of Justice of England and Wales refused the application to set aside the registration of the award under the Arbitration (International Investment Disputes) Act 1966 (which implements the ICSID Convention) to have it enforced in the United Kingdom, but granted a stay of enforcement pending determination of the proceedings in the EU General Court seeking annulment of the Commission’s decision. Mr Justice Blair also rejected the request of the Micula brothers that Romania should post security. The High Court’s decisions were appealed by the Micula brothers before the UK Court of Appeal. On 27 July 2018, the UK Court of Appeal upheld the High Court of Justice’s decision to stay the execution of the award pending the determination of the annulment proceedings before the General Court. However, it ordered Romania to post EUR 150 million in security – albeit without conditioning the stay on the posting of such a bond. The decisions of the High Court and the Court of Appeal, attempt to strike a balance between the obligation of the UK to enforce arbitral awards under the ICSID Convention and the country’s duty of loyal cooperation. Significantly, in the aftermath of the withdrawal of the UK from the EU the UK Supreme Court accepted the appeal brought by the Micula brothers and lifted the stay of enforcement. Considering that EU law still applies during the implementation period of the withdrawal agreement, the decision was based on the ground that the duty of sincere cooperation did not apply in this case on the ground that the UK had concluded the ICSID convention before its accession to the EU, and the enforcement of that convention thus sat outside of the jurisdiction of the CJEU by virtue of article 351 TFEU. Yet, the reference made by the UK Supreme Court to the unlikely possibility of infringement actions by the EU Commission arguably suggests a link between the decision and the termination of the country’s membership of the EU.
The loyalty of national Courts to the EU legal system has largely been tested also in Romania, Belgium, and Luxembourg. On 3 February 2017, the Bucharest Court of Appeal reportedly stayed the enforcement procedure started by the Micula brothers until the decision of the EU General Court on the state aid decision. It appears that on 26 January 2016, the Court of First Instance of Brussels also declined to enforce the Micula award in deference to the state aid determination of the Commission. An appeal against the decision is reportedly pending. Similarly, on 21 March 2018, the Supreme Court of Justice Grand Duchy of Luxembourg released “the State of Romania from the Award made against it”. It is unclear whether any appeal is currently pending against that decision before the Luxemburgish court of appeal. Some months later, on 23 January 2019, the Swedish Nacka District Court equally upheld the decision of the Swedish Enforcement Authority to refuse to enforce the 2013 ICSID award in favour of the two Romanian investors. For what is relevant here, the Swedish court notably motivated its decision on the ground of the duty of sincere cooperation that binds the domestic courts of the Member States to the EU Institutions, as mentioned by the CJEU in Achmea. In particular, the Naka District Court observed that the decision of the Commission that Romania’s tax cuts constituted illegal State aid, although under appeal, meant that the principle of sincere cooperation prohibited the enforcement of the award in Sweden, as that enforcement would “contribute to side-stepping” the action brought by the Commission. From another standpoint, the domestic courts of the Member States have not only showed their readiness to stand by their loyalty obligations vis-à-vis the Court and the other EU Institutions. They also have been very active in referring novel questions to the CJEU regarding the compatibility of awards of investment tribunals with EU law that ultimately help clarifying the scope of the Court’s decision in Achmea on the compatibility of ISDS tribunals with EU law. A particularly interesting case for the future of investment arbitration in the EU is developing before the Swedish courts: PL Holdings v Poland, a case arisen in the framework of the Luxembourg-Poland BIT and administered under the SCC arbitration rules. The arbitration tribunal in that case eventually awarded PL Holdings, the Luxembourgish investor, US$ 170 million in damages for indirect expropriation of its majority shares in a Polish bank. The award was almost entirely upheld by the Svea Court of Appeal on two main grounds: that Achmea cannot be read as in principle prohibiting all arbitrations between investors and EU Member States and that the validity of the arbitration agreement withstanding the case at issue can only be challenged if one of the parties expressly does so. Considering that, unlike in Achmea, Poland had not challenged the jurisdiction of the arbitration tribunal at the initial stage of the arbitration despite the clear time limitations under the SCC arbitration rules, the Court concluded that the Polish government had willingly consented to the arbitration. While upholding the award almost in its entirety, the Svea Court of Appeal exceptionally granted however leave to Poland to appeal its decision to the Supreme Court. As a result of the appeal, “[I]n order to avoid the risk of an incorrect interpretation of EU law” on 12 December 2019 the Swedish Supreme Court referred to the CJEU the interpretation on which the decision of the Court of Appeal of whether, by failing to raise the jurisdictional objection to arbitral tribunals’ jurisdiction in intra-EU BITs, Member States can still implicitly consent to arbitration. The question referred to the CJEU will be defining to delineate the scope of external autonomy: should the CJEU reject the interpretation put forward by the Svea Court of Appeal that Member States need to at least contest the jurisdiction of arbitral tribunals deciding on intra-EU disputes to have their previous international obligations set aside, the Court would effectively broaden the consequences of autonomy of the EU legal system in a way that is deeply at the odds with the identity of the EU as an actor open to dialogue with international courts. Such a conclusion would indeed not only unilaterally liberate Member States of their international pre-accession obligations towards international law, in violation of the VCLT. In so that Member States would not even have to put forward a defence against investment claims under intra-EU BITs, that liberation would also become an automatic and inevitable consequence of the intra-EU nature of the BIT at issue. In conclusion, the discussion above arguably supports an argument that more reliance and trust in the domestic courts of the Member States on the part of the Court is not only well-placed, but even beneficial to the political aim of the “ever closer Union.” On the one hand, it evidences that domestic courts have ‘interiorized’ their role of first point of delivery of EU law in the system of judicial protection of the EU and can effectively carry it out. On the other hand, it demonstrates that more reliance and trust in the domestic courts of the Member States strengthens the ties among EU and national courts, which, in cases like cases such as PL Holdings, are encouraged to critically interpret and apply the case law of the CJEU to the facts before them, stimulating a constructive dialogue among themselves, with the Court, and, indirectly, with investment tribunals. In that context, even reasoned disagreements have the positive outcome of stimulating a sense of participation in the same political project of bringing together the “peoples of Europe.” The Court should thus abandon the approach of ‘protection and prevention’ on the bases of which it has to date justified many declarations of incompatibilities of arbitral tribunals with EU law. Should those courts, after many years of European integration, suddenly refuse to exercise their powers to safeguard EU law, the Court would be dealing with different and profound constitutional issues, which would arguably render any arguments in terms of protection of the ‘external’ autonomy of the EU pointless.
4. handling conflict in accordance with the eu treaties
The previous section argued that issues of external autonomy with international dispute settlement mechanisms exercising outer scrutiny of the EU only arise in relation to conflicts in the ‘narrow’ sense, i.e., cases where that outer scrutiny creates a situation in which Member States or the EU Institutions may fulfil their obligations under international law only by thereby failing to fulfil another corresponding obligation under EU law. While the approach of openness towards outer scrutiny enshrined in the EU treaties and in external policy of the EU requires to adopt a narrow understanding of conflicts in the ‘narrow’ sense, once the existence of such conflict is ascertained, the issue arises of which primary law consideration – safeguard of the constitutional traits of the EU legal system or engagement with international dispute mechanisms exercising outer scrutiny over EU law – should prevail. The sections below put forward some suggestions on how to strike this balance.
4.1 Abandonment of the primacy claim and primary law rules
Implementing autonomy as ‘controlled openness’ firstly entails the abandonment on the part of the CJEU of the primacy claim discussed in chapter […], which leads to resolve conflicts in the ‘narrow’ sense by default in favour of EU law, to instead privileging the opposite principle, i.e., that the Court should in principle accept the existence and operation of outer scrutiny of international courts and tribunals and opt for the normative interpretation of EU law most in line with them. Prevalence of EU law or its system of judicial system at the expenses of the existence and operation of outer scrutiny of international courts and tribunals should instead be viewed as an exception limited to cases considered essential for the safeguard of the constitutional structure of the EU. As noted in chapter […], that is due to the fact that the EU treaties and the EU external policy require that the EU adopts an approach of principled openness towards engagement with international dispute settlement mechanisms. That is reflected in the internal hierarchy of norms of the EU, delineated by the Court itself, under which international agreements binding on the Union, including those that set up and regulate engagement of the EU with international dispute settlement mechanisms, are normally considered to be situated above secondary law and internal legislation. Therefore, secondary law must be interpreted as far as possible in a manner that is consistent with their terms. The same also applies to the rules of customary international law and other unwritten general sources of international law which govern the relationships among international actors. Only the nature of the conflicting rule of EU law as primary law justifies a declaration on the part of the Court of prevalence of EU law and its system of judicial protection at the expenses of the existence and functioning of international dispute settlement mechanisms. Therefore, when deciding whether that existence and functioning creates a conflict in the ‘narrow’ sense it is important to firstly ascertain the position, whether or primary or secondary law, of the allegedly conflicting norm(s) of EU law. While secondary norms will in principle have to pave the way to the obligation or norm arising from the operation of the dispute settlement mechanism, conflict with the founding treaties and other elements of primary law of the EU may be resolved in favour of EU norms. This conclusion emerged in the case law discussed above in section […] and was spelled out particularly clearly in Opinion 1/17, where the Court affirmed that the provisions of such international agreements may be interpreted by other courts of non-Member States which those agreements were concluded or that of the international courts or tribunals that established by such agreements provided “that the indispensable conditions for safeguarding the essential character of those powers are satisfied and, consequently, there is no adverse effect on the autonomy of the EU legal order.” The issue that arises immediately after is whether all rules of primary EU law may be deemed to prevail over the existence and operation of international dispute settlement mechanisms.
a. Public law rules of an essential natureThe approach of openness towards engagement with outer scrutiny of international dispute settlement mechanisms envisaged by the EU treaties and in the EU external policy suggests that not all rules of primary law may be considered as prevailing on international law at all times, purely by virtue of their nature. As it is the case with public policy, prevalence of EU law over international law should be triggered only by “some fundamental principle of justice, some prevalent conception of good morals, some deep-rooted tradition of the common weal.” Both the CJEU and the doctrine seemingly accept this premise. As it is evident by the language it uses, the Court tends to base its decisions on external autonomy on rules and principles of primary EU law that it deems particularly important. So, the Court refers to provisions such as articles 344 or 267 TFEU, or to fundamental rights, as the ‘very foundation’ or ‘essential elements’ of the EU legal, that cannot be compromised by international law. The Court also adopts the same approach when interpreting the scope of public policy in the national legal system of the Member States as a limit to the application of EU law: domestic public policy rules are normally recognizes by the CJEU where they protect a sufficiently important interest of society. Commentators discussing the role of public policy in the EU analogously concede that only few fundamental public law rules of EU law may prevail over competing rules of international law. Significantly, de Boer notes that “the only European rules that do deserve the status of overriding mandatory rules – and could be seen, therefore, as guidelines for the invocation of European public policy – are those that are primarily meant to protect European public interests. [T]he cases in which the public interests of the European Community are actually at stake will be relatively rare.”
4.2 Application of the fundamental public law rules of the EU
The identity of the EU as an entity that strives towards engagement with the outer scrutiny of international courts and tribunals arguably does not justify the default prevalence of primary EU law rules of a public nature over competing rules and obligations arising from that engagement. With the exception, to an extent, of Opinion 1/17, all the CJEU decisions discussed above to an extent may be regarded as upsetting the international-law friendly nature of the block and the role of international dispute settlement mechanisms in the EU treaties and in the EU external policy as a tool to reinforce and expand the values and interests of the Union abroad. Incompliance on the part of the EU with the rules and obligations deriving from outer scrutiny shall be regarded also in those cases as an exception. While the decision on whether and to what extent those essential rules of primary law should prevail rests with the CJEU, a number of general rules of application drawn from a large body of literature on public policy may guide the decision of the Court to that effect.
a. Seriousness of the harm
Autonomy as ‘controlled openness’ firstly suggests that prevalence of primary EU law rules of public policy over rules and obligations deriving from outer scrutiny may only be allowed where the application of the conflicting rules and obligation would create overall unbearable consequences for the constitutional structure of the EU. The decision of what may constitute an unbearable consequence for the constitutional structure of the EU rests, once again, with the CJEU. However, the following criteria may be considered. To begin with, incompliance should only be admissible when the conflict with the competing rules and obligations deriving from outer scrutiny is flagrant. It is indeed quite imaginable, that the application of a rule of such rules and obligations may in practice have limited practical consequences in the EU legal system. In that case, the presumption in favour of the application of the rule of international law should arguably stand. For instance, the admission by the Court in Gazprom that not all competition law rules qualify as ‘EU public policy’ rules, able as such to render unenforceable arbitration awards conflicting with them, as it also appeared to have been suggested in Eco Swiss, evidences that an array of practical impacts, not all equally damaging, may arise from the impingement of commercial awards in rules of competition law. Only the most flagrant and concrete among them may justify the disapplication of arbitral awards (or of decisions of other international courts or tribunals, for that matter). The same approach seems to have been followed by the national courts of the Member States, which have refrained from annulling arbitral awards on grounds of public policy when those awards did not ‘manifestly’ disregard EU law.
b. Proportionality of the harm
Arguably, a declaration of incompatibility of rules and obligations deriving from outer scrutiny with rules of primary EU law constituting ‘fundamental provisions of public law’ should also be based on a proportionality determination, aimed at weighting the harm caused by the application of the allegedly incompatible rule or obligation against the harm caused by its disapplication. The reasoning is that the interpretative exercise of the Court of deciding which one of the conflicting rules or obligation – deriving an international dispute settlement mechanism or EU law – to apply, requires balancing two sets of primary rules of EU law: primary rules of EU law that could be upset by the outer scrutiny of the dispute settlement mechanism, and primary rules entrenching principled openness towards that outer scrutiny. Therefore, in deciding that conflict, the Court should evaluate the prevalence of which set of considerations is likely to be, in the specific case at issue, the least problematic in relation to the constitutional framework of the EU. In Achmea, for example, the Court arguably did not balance the impact of its decision to declare intra-EU BITs tout court incompatible with EU law against the damage to the investors in pending or concluded investment treaty cases. There is no evidence in the judgement that, for instance, the CJEU considered the impact of its decision on the right to judicial protection of the investors in the form of the relevant time and resources already invested in investment arbitrations under intra-EU treaties, nor that many of them would be barred from starting another judicial action before national courts due to jurisdictional and procedural hurdles. Nor did the Court consider the position of those investors who had at the time already invested in Member States confiding in the special protection, in the form of substantive and procedural standards, of the intra-EU BITs.
c. Strength of the connection
The determination of a conflict with the rules and obligations deriving from the outer scrutiny of an international dispute settlement mechanism also requires an evaluation of the strength of the connection of the facts of the case at issue with the EU legal system: only a sufficiently strong connection with the EU legal system may justify the prevalence of EU law over the competing rules and obligations. That test resembles the ‘centre of gravity’ test carried out by the Court when identifying the predominant purpose of the measure in order to select the most appropriate legal basis for EU measures. For instance, one may argue that the CJEU did not correctly carry out this balancing exercise in Opinion 2/13 where it held that article 33 ECHR, providing for an inter-State dispute resolution mechanisms, may, in case of accession of the Union to the Convention, threatens the monopoly of application and interpretation of EU law of the Court under article 344 TFEU. It was indeed seen above that it remains unclear on what treaty basis the CJEU rules out that the EU Institutions may allow an external body to review the compatibility of the case law of the Court against an external comparator, the ECHR. Considering that the ECtHR is not called to assess the decisions of the CJEU against the EU treaties, that outer scrutiny certainly does not compromise the right of the Court to make its decisions and for that reason it does not impinge in the unity of EU law that article 344 TFEU protects. As an effect, the decision substantially seems to outlaw that any disputes among Member States may potentially be decided by the Strasbourg court. In so doing, the Court overlooked that the ‘horizontal’ nature of human rights, which touches upon a variety of different subject matters, means that a number of potential intra-EU disputes under article 33 ECHR may regard issues falling outside of the direct scope of EU law, and thus the jurisdiction of the CJEU. As Eeckhout also notes, in so doing, the Court also breaches a fundamental element of the system of law of the Council of Europe – i.e., the principle of equality of the contracting parties to the ECHR – in such a way that cannot be justified under EU law.
d. Hierarchy of values
The decision on whether to allow EU law and the EU system of judicial protection to prevail over conflicting rules and obligations arising from the outer scrutiny of international courts and tribunals also requires making considerations in terms of hierarchy of values of the EU legal order. The situation may well arise where, in practice, the application of a norm of international law overall safeguards a fundamental value of the EU better than the conflicting rule of primary EU law. Considering the ‘living’ nature of constitutional entities, that decision should be taken on the grounds of the status of the case law at the date of the court decision and strictly motivated on the grounds of the facts of the case. So, for instance, the evolution of the constitutional identity of the EU from a market-based to a value-based entity discussed in chapter […] should arguably lead the Court today to in principle privilege the application of the rules and obligations that overall better protects individuals’ rights over the one that privileges merely economic interests. Achmea is also an example of a case where the Court disregarded considerations in terms of balance of values safeguarded by the competing norms that it was called to apply. In that decision the CJEU effectively privileged the protection of a potential threat to the uniformity of EU law and to the systemic functioning of preliminary reference mechanism – i.e., that arbitral tribunals may bring litigation concerning EU law outside of the EU judicial system – over one of the actual and effective judicial protection of individuals, one of the pillars of the standards of the EU rule of law. The effect of Achmea was indeed an actual, retroactive deprivation for the investor of the protection afforded to them by the investor-state mechanism included in the intra-EU BIT through the unenforceability of an award already rendered by the international tribunal, which ultimately also deprived them of alternative routes of protection.
5.tailoring outcomes
In the exceptional circumstance that an actual conflict between an essential rule of primary EU law and rules or obligations arising out of outer scrutiny of international courts and tribunals have been ascertained, which also has such detrimental consequences for the EU legal system to render the application of those rules or obligations incompatible with the essence of the constitutional structure and values of the EU, that conflict may justify a declaration of incompatibility of that outer scrutiny with EU law on the part of the CJEU. It is however argued that autonomy as ‘controlled openness’ requires that the CJEU carefully tailors the language of such decisions to precisely define their scope and convincingly justify its conclusion.
5.1 ‘Priority’ over ‘validity’
A declaration of incompatibility of an international obligation of the EU responding to the requirements of autonomy as ‘controlled openness’ should arguably firstly be based on a reasoning in terms of ‘priority’ of one set of constitutional considerations over another, rather than of ‘validity.’ The reasoning of the CJEU should evidence that the ‘disapplied’ rule or obligation arising from outer scrutiny has not been rendered invalid by its decision. Rather, it will remain in the ‘background’ and influence the application of the rule of EU law to which priority has been given and may well be deemed prevalent in another case. An approach to conflict resolution in terms of ‘priority’ over ‘validity’ is in line with the approach of openness towards outer scrutiny of the EU treaties and EU external policy and sends a signal to all potentially concerned national and international courts that the disapplied rule or obligation should be kept into account in relation to all other issues that do not directly concern the subject-matter of the incompatibility.
5.2 Scoping the decision
Autonomy as ‘controlled openness’ also requires that the Court should carefully circumscribe the scope of its decisions to the specific facts of the case at issue. It is indeed quite possible that conflicts that on one occasion may be unavoidable, may be easily overcome in another. Arguably, only foreign rules in blatant violation of rules universally regarded as fundamental – one may use the extreme case of violation of jus cogens rules – may automatically be held inapplicable regardless of the circumstances and results of their application. Outside of these very rare cases, declarations of incompatibility should thus only represent the outcome of a specific analysis on the part of the CJEU of the content and application of the conflicting rules and obligations arising from outer scrutiny on a case-by-case basis. Among the examples discussed above, the most problematic are probably Achmea and Eco Swiss: in both cases the Court concluded with general statements that, in the former case, “Articles 267 and 344 TFEU must be interpreted as precluding a provision in an international agreement concluded between Member States, such as Article 8 of the BIT […]” and, in the latter case that “a national court to which application is made for annulment of an arbitration award must grant that application if it considers that the award in question is in fact contrary to Article 85 EC […].” The result of this generalisation on the part of the Court is a potential broadening of the scope of the conflicts in its ‘narrow’ sense to all arbitration clauses included in intra-EU BITs (Achmea) and to all issues of competition law (Eco Swiss). The extension of the scope of conflicts in the ‘narrow’ sense beyond the specific case in which they arise also sits uncomfortably with the CJEU’s constitutional decision-making role as an ‘adjudicator.’ In so that it supports declarations of full incompatibility with EU law of the outer scrutiny of any given international dispute settlement mechanism to which the EU or its Member States are participating, that extension arguably comes close to conferring to the Court a ‘treaty making’ function of terminating that international commitment, which more properly belongs to other EU Institutions or to the Member States. From this standpoint, the importance that the Court limits itself to decide on the specific circumstances of each case can be framed as a treaty requirement to preserve the prescribed institutional balance. The importance of carefully scoping the impact of these incompatibility decisions of the Court also arguably leaves little space for a priori declarations of general incompatibilities between the EU legal system and outer scrutiny of a certain international dispute resolution mechanism. Along that line Advocate General Wathelet in its opinion in Achmea specifically rejected a ‘systemic’ incompatibility between the international investment arbitration in intra-EU BITs and EU law. The AG highlighted the importance that the Court evaluates in its decisions the “real threat” that the intra-EU investment arbitration poses to the uniformity and effectiveness of EU law, which he finds “greatly exaggerated” given the low rate of success of investors in intra-EU arbitral proceedings, the fact that arbitral tribunals have to a large extent allowed the Commission to intervene in arbitrations, and that arbitral tribunals have normally not been required to review the validity of acts of the Union or the compatibility of acts of the Member States with EU law. Beyond the merits of the reasoning of AG Wathelet, its value resides foremost in the plead that he makes to the Court to avoid a priori declarations of systemic incompatibilities between EU law and international dispute settlement mechanisms. Similarly, in her advice delivered in the framework of Opinion 2/13 on the accession of the EU to the ECHR, Advocate General Kokott found against a general rule on “the inadmissibility of inter–State cases before the ECtHR and [on] the precedence of Article 344 TFEU as a prerequisite for the compatibility of the proposed accession agreement with EU primary law,” noting that many longstanding international agreements of the EU which do not contain that rule have never raised issues of incompatibility in practice. Eeckhout notes that those agreements include, for instance, the DSU of the WTO, to which all Member States are party. The author notes that, despite the large overlap between the exclusive jurisdiction of the WTO’s DSU established under article 23 of that agreement and the jurisdiction of the CJEU under article 207 TFEU, no intra-EU cases have ever been brought before a WTO panel, nor a suggestion in that sense was ever paved.
5.3 Dialogue
Finally, autonomy as ‘controlled openness’ arguably also requires that the exceptional cases in which the CJEU finds the incompatibility of the outer scrutiny of international dispute settlement mechanisms with EU law should attempt to engage in a dialogue with international law. As noted in chapter […], that need for dialogue is part of the identity itself of the EU. As it has been argued at lengths in this Ph.D., dialogue with national and international law indeed supports the definition of the specific contours of EU law not only through agreement but even through reasoned divergence, which aids the safeguard and spreading of EU values beyond its confines. As Eeckhout points out, judicial dialogue is also intrinsic to the DNA of the CJEU itself both in its function as the constitutional court of the EU and the international dispute resolution mechanism of the EU: on the one hand, article 267 TFEU makes of dialogue a key element of the constitutional identity of the Court. On the other hand, articles 6, 21(1), 21(2) h) TEU arguably enshrine into primary law a requirement that the Court collaborates with its international counterparts. Informal dialogue has also been particularly influential in the EU legal system, as it is evidenced by the fact that the system of protection of fundamental rights in the EU is the result of ‘pressures’ on the CJEU by the Italian and German constitutional courts. The Court has on various occasions proven to being open to building some dialogue with other courts and tribunals, particularly the ECtHR, as it emerges in Matthews and Spain v United Kingdom. In the former decision, the Strasbourg court confirmed the right of the citizens of Gibraltar to vote in elections of the European Parliament. In the latter decision, Luxembourg dismissed the claim brought by the Kingdom of Spain against the steps taken by the United Kingdom to comply with the decision of the ECtHR to enable the Gibraltar electorate to vote in elections to the European Parliament. Bronckers also notes how the CJEU occasionally also entertains a “muted dialogue” with the WTO’s DSU. For instance, in the Boneless Chicken Cut Tariff case the CJEU changed its precedent case law and declared the invalidity of Commission regulation no. 1223/2002 on the basis that the Institution had altered the heading of the tariff classification of boneless chicken cuts laid down by the World Customs Organization (WCO), as interpreted by the WTO Appellate Body in the EC-Chicken Cuts case. The Court concluded that, overstepping the competence conferred to the WCO, the Commission had exceeded the discretion conferred upon it by the Council. Interestingly, the decision of the CJEU does not expressly mention the position of the WTO Appellate Body on the litigated heading, despite that decision having been expressly mentioned during the proceedings. Many definitions of ‘dialogue’ exist in legal theory, which essentially depend on how one theorises the relationship among legal systems. However, from the standpoint of this section of ‘tailoring the outcomes’ of the exceptional decisions of incompatibility of international rules or mechanisms with EU law, dialogue will be understood foremost as ensuring that the international dispute settlement involved in the decision of incompatibility of the CJEU must be able to understand the rationale of that decision. The argument is that only that understanding will allow them to ‘respond’ to the position taken by the Court – through acceptance or conflict – triggering a series of political and legal exchanges with the EU that comply with an approach of ‘controlled openness’ in line with the EU treaties and policy approach of the EU to the external relations.
a. Reframe decisions in international law
As discussed, it is essential to the abandonment of the primacy claim of the CJEU over outer scrutiny of international dispute settlement mechanisms that the Court takes off its ‘hat’ as the constitutional court and takes the role of the international dispute resolution mechanism of the EU when deciding issues having broad international law consequences. Among other things, in practice this means that the Court should in principle refrain from reasoning its decisions of incompatibility of international rules purely on an analysis of EU law rules and principles. Rather, the CJEU should also consider the general context of – and impact on – the rules of international rules and obligations arising from the outer scrutiny of an international dispute settlement mechanism that that it intends to disapply. That was for instance evident in Achmea, where the Court overlooked the grounds and consequences under international law of the declaration of incompatibility of the dispute resolution clause declared incompatible with EU law. Instead of simply assuming that EU law ‘replaced’ the previous international obligations of the Member States and ‘wishing away’ the previous international obligations of the Member States, the Court should have explained whether article 8 of the Dutch-Slovak BIT had merely been disapplied in that specific instance on the grounds of public policy, or it was to be considered terminated under the VCLT. In this latter case, considering that the EU is bound by the general principles of international treaty law codified in the VCLT, the CJEU should probably also have discussed on what grounds that termination had happened – whether on the basis of article 59 VCLT (termination or suspension of the operation of a treaty implied by conclusion of a later treaty) or article 30 VCLT (lex posterior). The Court could have arguably also discussed how Achmea fits, if at all, with the principles of pacta sunt servanda under article 26 VCLT, and that a party (the Member States) may not invoke the provisions of its internal law as justification for its failure to perform a treaty under article 27 VCLT. Instead, the decision of the Court in Achmea left investors, Member States, and international tribunals alike puzzled on the fate of intra-EU BITs. The lack of clarity on the effects of Achmea triggered a number of uncoordinated and conflicting attempts to reframe that decision within the framework of the law of the EU treaties that ultimately further damage not only the judicial protection of individuals, but, more generally, the rule of law.
b. Reasoned divergence
It was argued above that it is only physiological that a certain legal situation may be governed by competing sets of legal rules and be differently interpreted in the context of different sub-systems of law. The CJEU itself at times applies norms of international law, even key provisions of international law, such as the rules of the VCLT, in a way that deviates somewhat from the accepted approach in public international law. In these situations, autonomy as ‘controlled openness’ arguably requires that, when the Court disapplies, or differently applies, rules or obligations arising from outer scrutiny, the Court clarifies the existence and rationale for its decision. Reasoned divergence by the CJEU complies with the constitutional identity and policy approach of the EU towards engagement with outer scrutiny in so that it turns a pledge to uphold the rule of law internationally from an empty resolution into an active commitment. The case in Achmea offers once again a useful example of (lack of) reasoned divergence. After analysing the international law framework of its decision as discussed in the previous chapter, the Court could have entered into a discussion on why the ‘specific nature’ of the EU and the principle of loyalty justify – in its eyes – the incompliance with international rules due to EU public policy. While that explanation would probably have probably still left many unsatisfied with the approach of ‘European exceptionalism’ of the CJEU, it would have also had the positive outcome of allowing ‘reasoned’ reactions. From a standpoint of judicial comity, that more open approach would have also had the additional positive outcome of inviting judicial cooperation with investment tribunals, bringing closer together the now very separate planes of national and international law.
6. The future of ISDS: reform and choice
Before finally moving to the conclusions, the focus of this Ph.D. on ISDS arbitration begs the question of what approach should the Court take to that dispute settlement mechanism. It should firstly be clarified that the arguments developed in this Ph.D. in support of ISDS arbitration don’t aim to portrait that mechanism as “ideal.” The ‘backlash’ against ISDS arbitration that has given rise to the reform process in which the EU is also currently involved certainly has its merits. Indeed, ISDS arbitration has been often used as a sword for rich investors to pursue their interest at the detriment of developing countries. Admittedly, when dealing with human rights, of non-state entities involved in investment disputes – parties or non-parties – investment tribunals have often been adopted a position of restraint. ‘Double-hatting’ and diversity in the arbitration community, just to name two, are well-known issues to the arbitration community. Yet, ISDS arbitration is, after all, what states make of it. Most of the issues discussed above constitute the result of choices made by states in the drafting of investment agreements. For instance, the position of restraint of arbitration tribunals on dealing with human rights issues results from the limited jurisdiction that arbitrators were conferred by investment treaties and was aimed at avoiding criticisms of judicial activism of the contracting parties. Similarly, the issues of double-hatting and diversity arguably are the legacy of ISDS as an ‘elitist’ system build by western countries, which is today evidently outdated and requires reform. These criticism arguably don’t take away the merits of the entire ISDS regime as an international dispute settlement mechanism that can effectively uphold the rule of law internationally, and certainly the one that, in doing so, most clearly put investors and states on a position of (almost) equal footing. Reforms have been carried out, and many others are needed, to correct many of the issues with ISDS arbitration. These include enhanced transparency, mechanisms to reduce costs, including expedite arbitrations, mechanisms for the early dismissal of frivolous claims, mandatory disclosures on third party funding, stricter rules for the independence and impartiality of arbitrators. Notably, today greater attention to human rights in investment disputes is made possible through a variety of reforms in treaty making, which include the explicit incorporation of human rights and sustainable development considerations into the treaty text, the explicit protection of States’ rights to regulate, possibility of claims against the investor in their home State, and, in some cases, even the explicit imposition of obligations on investors. Arguably, with these improvements, ISDS arbitration certainly deserves a place among the possible avenue to pursue the rule of law in international investment law. It is indeed maintained that, when it comes to ensuring protection of rights, especially individual rights, more is better than less. Admittedly, whether that will be the case for the EU is a matter for the Commission and the Council, which unwisely appear determined to get rid of it, rather than for the Court. Yet, as it is likely that the CJEU will be involved in decisions on various aspects relating to the compatibility of ISDS with EU law for still a long time to come –in doing so, it should arguably consider the important function and effectiveness of these mechanisms in international law. Not least for the reason of not contributing to the undoing of 50 years of advancement of the protection of individual’s rights in international law.
7. conclusions
This chapter argued that the identity of the EU as an entity that has entrenched in its own DNA the importance of outer scrutiny of international courts and tribunals to uphold the rule of law internationally should lead the Court to reframing its interpretation of external autonomy as ‘primacy’ into one of ‘controlled openness.’ Section […] observed that autonomy as ‘controlled openness’ does not demand that the CJEU shies away from its constitutional role of safeguarding the essential traits of the EU legal system from potentially detrimental impacts of the outer scrutiny of international courts and tribunals: the Court indeed retains its position as the ultimate bastion of protection of the essential functioning mechanisms of the Union and its system of values. As Kleinheisterkamp observes, “when framed and applied correctly, indeed, [those mechanisms and values constitute] the lois de police that enforce a minimum of respect for national and EU public policies.” As much as they may conflict with the commitment to a broad engagement with international courts and tribunals exercising outer scrutiny over the EU, those mechanisms and values perform an essential function in the EU legal order, which finds the condition for its own survival and essential guidance for its development within them. They thus need to be safeguarded. However, autonomy as ‘controlled openness’ does demand that the CJEU adopts an approach of enhanced dialogue with – and openness towards – international law. That shift of approach primarily requires the abandonment of the ‘sovereign’ claim that EU law and its system of judicial protection in principle prevail over the operativity of international dispute settlement mechanisms exercising outer scrutiny over the EU, and instead embrace the position of principled acceptance of that scrutiny and the compromises to the sovereign powers of the EU that it entails. This chapter attempted to bridge the theoretical account of autonomy as ‘controlled openness’ with its understanding by the CJEU by putting forward a number of suggestions on how the Court may in practice implement this shift. It did so by, firstly, proposing in section […] to reduce the instances of actual conflict between EU law and international dispute settlement mechanisms to a handful of very specific instances where those mechanisms impose rules or obligations that are irreconcilable with obligations under EU law (conflict in the ‘narrow’ sense) of the EU or its Member States. Conflict in the narrow sense, it was argued, should be read as an exceptional instance, limited to those instances where no other outcome is possible, and which present a connection of a certain degree with the present time, i.e., taking into account the state of the law at the date of the court decision without assuming how that law will evolve or be interpreted in the future. Section […] then moved on to discuss possible instruments to deal with conflicts in the ‘narrow’ sense between EU law and international dispute settlement mechanisms. The suggestions were put forward that only conflicts in the narrow sense and that concern essential rules of public primary EU law may exceptionally justify incompliance with rules or obligations arising from the outer scrutiny of international courts and tribunals exercising outer scrutiny over EU law. Such incompliance should also be limited to cases in which (i) those rules and obligations would, if applied cause some serious harm to the EU legal system; (ii) disapplication of the rules of obligations is not disproportionate to the harm caused to the EU legal system by the violation of international law; and (iii) where the rule or obligation does not have the ultimate effect of safeguarding values and rights of the EU hierarchically superior to those safeguarded by the essential interest that the CJEU is trying to protect. Section […] concluded that, even when all these considerations are satisfied and the Court decides to proceed with a disapplication of rules or obligations arising from the outer scrutiny of an international court or tribunal, it should carefully limit the impact of its decision through three main instruments. Firstly, by clarifying that the relevant rule or obligation is not ‘invalid,’ but rather continues to exist in the background of the dispute despite not having been given priority in the specific case at issue. Secondly, by carefully defining the scope of its decision to the specific facts of the case. Thirdly, by the use of language, ensuring that its decisions also take into account the relevant rules or obligations that the CJEU is disapplying, and carefully reasoning divergence from the international dispute settlement mechanism. It was argued that these last two tools allow all the relevant national and international courts potentially involved in the dispute to understand the decision of the Court and tailor their response to it triggering an important dialogue of which the rule of law as a whole will benefit. Finally, section […] concluded with some thoughts on what approach should the Court take to ISDS arbitration in the future. argued that ISDS arbitration is certainly a perfectible dispute settlement mechanism, which however has marked one of the greatest advancements for the rights of individuals in international law in the last 50 years. While the choice of whether it should be retained in the EU as one possible avenue for investors to uphold their rights under investment agreements lays with the Commission and the Council, the CJEU will certainly be called many more times to decide on the compatibility between EU law and ISDS arbitration. Section […] concluded with the wish that the CJEU will, on those occasions, adopt a different approach to autonomy and recognize not only the pitfalls, but also the merit of that dispute settlement mechanism.
CHAPTER VI – CONCLUSIONS OF THE PH.D.
This Ph.D. addressed the understanding of the Court of ‘autonomy of EU law’ in relation to the outer scrutiny of international dispute settlement mechanisms, and how that understanding fits with the identity of the EU, delineated in the EU treaties and in the EU external policy, as a protector and promoter of the rule of law in international law. This final chapter of the Ph.D. aims to summate the conclusions of the thesis with some final reflections; identify some of its main challenges and limitations; and posit what could be next. Starting with the first point, three are the main conclusions of this thesis. Firstly, autonomy is broadly understood by the Court as a ‘doctrine of constitutional interactions,’ i.e., an interpretation of the Court of what the EU treaties provide at one specific point in time is necessary for the Union to fulfil the two main conditions for its autonomous existence of self-determination and independence. As a doctrine, autonomy also intrinsically provides the Court with large interpretative discretion, larger than the one the Court possesses in relation to other legal instruments of EU law, such as general principles. That arguably explains why the Court tends to expressly mention the term ‘autonomy’ in its most controversial decisions, as a tool of legitimisation. However, as a doctrine, autonomy can also only be developed within the constitutional framework of the EU. That framework enshrines the centrality of the protection of individuals rights and is based on pluralism and exchanges with national and international law, as the natural corollary of a decentralised and horizontal legal order with a specific history and own aims. Thus, autonomy cannot be interpreted in a manner that prejudice the protection of individuals, the identity of the EU, or dialogue with others. Secondly, the CJEU appears however understand the ‘external’ autonomy of EU law in relation to the outer scrutiny of international in a progressively divergent manner compared the constitutional provisions of the EU treaties and EU external policy, with serious repercussions on the legitimacy of the EU as a whole. It was seen that international courts and tribunals today are powerful tools to uphold the rule of law internationally and perform a function that goes well beyond dispute settlement, becoming instruments of global governance. That is particularly true for ISDS arbitration. Yet, engagement with them comes at the price of the cession of some sovereignty. While the EU treaties and the EU external policy today accept that compromise and embrace participation in international dispute resolution mechanisms as an essential element of the identity of the EU and for the achievement of its aims in the international relations, the Court seems more reluctant to do so. By adopting a very high degree of scrutiny of the compatibility with EU law of the outer scrutiny of international courts and tribunals, the Court instead implements the doctrine of autonomy in a way that closely recalls the adopt of ‘primacy’ traditionally characterising the internal relations. Among other things, it was noted that the Court seems to unduly measure the compatibility of international courts and tribunals with EU law based on the EU-specific substantive standards encompassed by the primary law right to judicial protection. A right that was developed as an expression of primacy and loyal cooperation of national courts in the new value-based framework of the treaties post-Lisbon, rather than in relation to the minimum standards of enforceability of EU law originally required to ensure the autonomous existence of the EU. This approach raises issues in terms of legitimacy and effectiveness of EU action in the external relations, and risks undermining its capacity to live up to its ideological blueprint of creating the “ever closer union” for various reasons. Not only does it disregard the very different circumstances of the external relations compared to the dynamics internal to the Union, and particularly that equally ordinated position of international courts and tribunals and the fact that they operate in separate legal frameworks that are not meant to converge. By conflicting with the constitutional nature and essential traits of autonomy discussed above, the approach of the Court also creates a number of issues in terms of internal constitutional structure, legitimacy, and overall role of the EU in the international relations. Specific issues relating to autonomy as primacy were also found in relation specifically to ISDS arbitration: it was argued that an approach of the Court of autonomy as primacy ultimately supports an unfounded idea of superiority of courts over arbitral tribunals and is liable to effectively deprive individuals of instruments for the protection of their rights available outside of the EU legal system, in violation of the constitutional identity of the EU. For all these reasons, this Ph.D. suggested that the CJEU should adopt a new paradigm of autonomy, more compatible with the identity of the EU as set out in the EU treaties. The thesis thus set up the coordinates of this new paradigm, defined as ‘controlled openness,’ and puts forward some suggestions as to how to implement it in practice. These include an invitation enhanced dialogue between the CJEU and international courts and tribunals, as well as the observation that ISDS arbitration, if accurately designed at treaty level, possess the potential to further many of the policy objectives of the EU and should not be discarded as one among many options for the effective protection of investors’ rights. The development of this paradigm of autonomy as controlled openness wanted to realise the ultimate aim of this Ph.D. to attempt to constructively contributing to the understanding on autonomy of EU law. It tried to do that by not only evidencing the pitfalls of the vagueness of the concept of autonomy and of the current ‘negative’ approach of the Court to outer scrutiny of international dispute settlement mechanisms, but also by putting forward practical proposals to overcome them in a way that may effectively contribute to the European project of the “ever closer union.” By defining a new paradigm of autonomy through which the CJEU can both preserve the essential characteristics of the EU and adopt a principled approach of openness towards international courts and tribunals that responds to its identity, this thesis intended to overcome the Court’s understanding of ‘autonomy of EU law’ as a constraining element for the development of international relations compatible with the EU treaties. Autonomy can instead become a tool to fully unleash the potential of the Union as a promoter of high standards of rule of law and judicial protection, turning outer scrutiny from a ‘threat’ to an instrument to safeguard, and even advance, the values of the EU of rule of law and judicial protection. So reframed, autonomy will ultimately contribute to the legitimacy of Union by making the EU Institutions fully able to realise the ideological blueprint of the EU treaties. It is suggestive to compare the proposed development of the Court’s understanding of autonomy to the relationship with otherness underlying the construction of the constitutional identity elaborated by Rosenfeld in his work ‘The Identity of the Constitutional Subject: Selfhood, Citizenship, Culture, and Community.’ The author reads the existence of interactions with ‘others’ as an essential element of the development of the constitutional identity of subjects – their ‘autonomy.’ Specifically, Rosenfeld describes the development of the constitutional identity as a constant pursuit of self-achievement that requires the construction of a relationship with subjects external to the constitutional subject. Particularly relevant for this thesis is the reconstruction made by Rosenfeld of how the development of the constitutional identity influences that relationship. The author makes in that regard an interesting parallel with the three phases of the Hegelian Phenomenology of Spirit: conflict, domination, and recognition. The first phase, conflict, goes through the affirmation of the self through negation of dependence from others. The second phase, domination, is acceptance that the self depends on others, while still maintaining that relation with those external. The third phase, recognition, completes and overtakes the first two, and it is discovery that both being recognized by others and the existence of mutual differences form an integral part of the self. Along this line, one may argue that, while the approach of the Court to autonomy as ‘primacy’ over international law could be justified in the origin on the grounds of the protection of the creation of an EU-specific identity, separate from national or international law, that approach is no longer reasonable or useful for a more mature Union, that confidently displays its unique attributes without a need to negate those of others. On the opposite, in a key to further advance the European project, the Court should become less formalistic in its approach to international law and recognize that it is in the work with, rather than against, others that lays the key to a more effective pursuit of the ideological mandate of the “ever closer Union.” In terms of challenges found in the development of this work, the main difficulty was perhaps the selection of the cases analysed above. As discussed in the methodology section, in the face of a very broad understanding of the Court of autonomy as a ‘doctrine of constitutional interactions,’ it was quite difficult to narrow down the pool of cases on the basis of which to assess the approach of the Court to autonomy and outer scrutiny over EU law. When discussing the specific interactions between autonomy and ISDS arbitration it was also difficult to define the limits of the discussion on the compatibility of the reform being carried out by the Commission in relation to the ICS and the MIC. This ties in with the limitations of this work. This thesis only briefly looked at the compatibility of the ICS and the MIC with the autonomy of EU law. It further did not engage with a discussion of that potential compatibility with other forms of arbitration, such as sports arbitration. This thesis did not assess the broader understanding of the Court of external autonomy as the constitutional interactions underpinning the relationship between the EU and international law as a whole, such in relation to the general principles of international law, participation of the EU in international organisations, etc. Nor did it assess the understanding of the CJEU of internal autonomy, as the constitutional relations among Member States and EU Institutions. This Ph.D. also could not engage with the topic of how autonomy is being understood and used by other EU Institutions which, as mentioned in the introductory chapter, are increasingly making use of that concept in their own interactions with international law. Further, as also discussed above, the proposals set out to implement the new paradigm of ‘controlled openness’ are only indicative and do not attempt to cover the universe of possible choices that the CJEU faces when undertaking the complex task of deciding on the constitutional interactions between EU and international law.
The challenges and limitations just mentioned arguably also frame the scope of what could be next. It will firstly be interesting to see whether and how in practice the ICS and the MIC, if they will ever become operable, will effectively be less intrusive of the autonomy of the EU than ISDS arbitration. Similarly, analysis will hopefully be carried out on whether they will maintain at least the same level of protection for investors’ rights that is now granted by ISDS arbitration. Some space for further research exists regarding the understanding of the Court of autonomy in relation to other areas of the external relations. Arguably, the described increased importance of international law in the EU treaties as a tool to fulfil the ideological blueprint of the Union is here to stay. It thus seems relevant to expand the theoretical framework construed in this thesis beyond its narrow scope. Future research could also address the relevant question of how internal and external autonomy interact among themselves, and whether their development follows similar evolutionary dynamics. That research seems relevant because, as this thesis hopefully managed to evidence, the Court’s understanding of autonomy remains essential to protect, strengthen, and realise the foundational values of the EU treaties and so guarantee the best possible future for the Union.
What Makes Us Unique
- 24/7 Customer Support
- 100% Customer Satisfaction
- No Privacy Violation
- Quick Services
- Subject Experts
