EU Membership: Formal and Substantive Dimensions

Abstract

Membership is central to the EU, as it is to any other international organisation. Withdrawal has assumed centre-stage through Brexit. While there is literature that is relevant to membership, most notably through academic discourse on differentiated integration, there is little more general inquiry concerning membership, the concept of which has importance and implications over and beyond more particular avenues of scholarship. This article examines the formal and substantive dimensions of membership and withdrawal in the EU.

I. INTRODUCTION

The concept of membership and its antonym withdrawal are central to any international organisation, and this is equally true for the EU. Accession speaks to choice and entry, withdrawal to voice and exit. Membership undoubtedly features in discussion of many aspects of EU law. This is unsurprising, since such analysis necessarily involves elaboration of the rights and obligations that flow from particular Treaty provisions, or the legal acts based thereon. It is, nonetheless, helpful to stand back and consider more generally the conception of membership that underpins the EU, since it reveals conceptual and normative assumptions that are not self-evident when the focus is solely on consideration of the duties entailed by a particular Treaty provision.

Part II addresses the formal dimension of accession and withdrawal, regulated by Articles 49 and 50 of the Treaty on European Union (‘TEU’), respectively. Part III considers the substantive dimension of membership in relation to Member States. There is analysis of the different types of primary and secondary obligations contained in the Treaty, and their interaction is exemplified through consideration of the rule of law problem and the use made of Article 19 of the Treaty on the Functioning of the European Union (‘TFEU’), and through the way in which Article 4(3) TEU has been used to reinforce Member State obligations. The focus in Part IV shifts to the substantive dimension of membership so far as it pertains to individuals, through direct effect and citizenship. Part V considers the substantive component of membership in the light of differentiated integration. There is discussion of the ways in which membership facilitates such differentiation, and also of the decisional, functional, and normative constraints that it places thereon.

The substantive facet of withdrawal is analysed in Part VI, in which it will be seen that Article 50, as interpreted by the Court of Justice of the European Union (‘CJEU’), reflects in part state sovereignty and voluntarism, and in part bilateralism and bifurcation. The conclusion in Part VII considers three more general features of membership as revealed by the preceding discussion, which concern the development of membership obligations across time, the fact that the constituent Treaties embody an incomplete agreement, and that membership requires clarity when thinking about the ascription of responsibility.

II. FORMAL DIMENSION: ACCESSION AND WITHDRAWAL

A. Accession

We begin with the formal dimensions of membership, as manifest in the rules concerning accession and withdrawal. It is readily apparent that the formal requirements are underpinned by normative precepts and values. This is evident in relation to Article 49 and Article 50 TEU. Article 49 is the central Treaty provision concerning accession.

Any European State which respects the values referred to in Article 2 and is committed to promoting them may apply to become a member of the Union. The European Parliament and national Parliaments shall be notified of this application. The applicant State shall address its application to the Council, which shall act unanimously after consulting the Commission and after receiving the consent of the European Parliament, which shall act by a majority of its component members. The conditions of eligibility agreed upon by the European Council shall be taken into account.

The conditions of admission and the adjustments to the Treaties on which the Union is founded, which such admission entails, shall be the subject of an agreement between the Member States and the applicant State. This agreement shall be submitted for ratification by all the contracting States in accordance with their respective constitutional requirements.

The anatomy of Article 49 TEU is revealing. It stipulates that unanimity in the Council is required for a Member State to join, plus consent of the European Parliament (‘EP’). The EU is founded on the twin constructs of Member State representation in the Council, and citizen representation in the EP.¹ It is therefore unsurprising that assent must be forthcoming from both the Council and EP. Unanimity in the Council betokens the centrality of accession to the functioning of the EU, hence the need for all existing Member States to be on board before a new state joins the club. The fact that the agreement between the acceding state and the EU must be ratified in accord with the constitutional requirements of each Member State attests to the importance of new membership, and that the legitimacy thereof is dependent on consent at the state level, not merely through the instrumentality of voting in EU institutions.

The Commission and the European Council are both formally involved, notwithstanding the fact that neither votes. Thus, the Commission is consulted by the Council before it votes, thereby reflecting the fact that it is the Commission that conducts the checks to ensure that the aspirant state conforms to the acquis communautaire and other conditions for accession. The statement in Article 49 to the effect that the conditions for eligibility set by the European Council ‘must be taken into account’ underplays the significance of this institutional actor. The reality is that the European Council determines the overall conditions for eligibility and determines also whether the EU should move ahead with accession of any particular state.

The success of any application will be dependent, inter alia, on whether the state is willing and able to comply with the acquis communautaire, and it will perforce be bound by Treaty obligations if the application is successful. Compliance with EU values listed in Article 2 TEU is a condition precedent to be considered for membership. This includes the Copenhagen criteria: a free market economy, stable democracy, compliance with the rule of law, transposition of EU legislation into national law, and acceptance of the euro.² Compliance with the Copenhagen criteria is a necessary condition for membership. It is not sufficient. Article 49 states that a state that complies with Article 2 TEU may apply to become a Member State, but there is no guarantee that it will be accepted. There has to be political acceptance of the new state, and the EU emphasises that accession must not ‘overstretch the capacity of the Union to integrate the new member’.³

The initial hurdle for states that aspire to EU membership is to be accepted as candidate states, in which capacity they become eligible for EU pre-accession financial assistance designed to ‘strengthen democratic institutions and the rule of law, reform the judiciary and public administration, respect fundamental rights and promote gender equality, tolerance, social inclusion and non-discrimination’.⁴ The Commission will, moreover, scrutinise closely whether the candidate state meets the acquis communautaire, which is for these purposes divided into 35 chapters.⁵

B. Withdrawal

The formal criteria for withdrawal are in Article 50 TEU. It was dissected on multiple occasions during Brexit, but it is nonetheless important to set it out here, since it facilitates the subsequent analysis:

1. 1. Any Member State may decide to withdraw from the Union in accordance with its own constitutional requirements.

2. 2. A Member State which decides to withdraw shall notify the European Council of its intention. In the light of the guidelines provided by the European Council, the Union shall negotiate and conclude an agreement with that State, setting out the arrangements for its withdrawal, taking account of the framework for its future relationship with the Union. That agreement shall be negotiated in accordance with Article 218(3) of the Treaty on the Functioning of the European Union. It shall be concluded on behalf of the Union by the Council, acting by a qualified majority, after obtaining the consent of the European Parliament.

3. 3. The Treaties shall cease to apply to the State in question from the date of entry into force of the withdrawal agreement or, failing that, two years after the notification referred to in paragraph 2, unless the European Council, in agreement with the Member State concerned, unanimously decides to extend this period.

4. 4. For the purposes of paragraphs 2 and 3, the member of the European Council or of the Council representing the withdrawing Member State shall not participate in the discussions of the European Council or Council or in decisions concerning it. A qualified majority shall be defined in accordance with Article 238(3)(b) of the Treaty on the Functioning of the European Union.

5. 5. If a State which has withdrawn from the Union asks to rejoin, its request shall be subject to the procedure referred to in Article 49.

We are concerned here with the formal dimension of Article 50, the substantive aspects thereof will be considered below. There are certain noteworthy features of Article 50 in this regard. Thus, it is for the Member State that seeks to withdraw to trigger Article 50 in accord with its constitutional requirements. If a withdrawal agreement is concluded, it only needs approval by qualified majority, not unanimity.

Article 50 is also formally predicated on bifurcation of the withdrawal agreement and the agreement on future trade relations, assuming that the latter is concluded. This bifurcation is both rational and problematic in equal measure. It is rational because an agreement on future trade relations may take a number of years to conclude. It would be unacceptable from the perspective of the withdrawing state to be forced to remain in the EU for this period of time. It would be equally unacceptable and disruptive from the EU's perspective for a state that has made clear that it wishes to leave then to remain within the EU for two or three years.

The bifurcation is nonetheless problematic, since it almost certainly means that the consequences of exit will be wholly or partially blind for the withdrawing state at the time when it completes withdrawal. The people will vote whether to leave or remain in the EU with scant if any idea as to the nature of the future trade relationship with the EU thereafter. The legislature of the withdrawing state will assess the withdrawal agreement without knowing the content of the trade deal that will define relations between the withdrawing state and the EU. This is so even though the content of the future trade deal will be more important for the country concerned and its citizens than will the withdrawal agreement. The withdrawing state may have thought through its preferences concerning the future trade relationship before deciding to invoke Article 50, but it will probably not have done so in any great detail. The preferences of the withdrawing state in this regard may change as a result of shifting political fortunes at domestic level, and there is, in any event, no assuredness that those preferences will be met, since they may not be accepted by the EU.

C. Accession and Withdrawal: Compare and Contrast

Accession and withdrawal are different, axiomatically so. It is nonetheless interesting to compare and contrast the formal dimensions of membership concerning accession and withdrawal. This comparison can profitably be undertaken in three respects: the agreement to get in/out, the trade terms thereafter, and the voting rules.

When viewed against these criteria, accession is simpler and is akin to a ‘prix fixe’ deal. The accession agreement is predicated on the default position that the acceding state complies with the entire acquis as a pre-condition of membership. It is possible for the acceding state to negotiate opt-outs or qualifications to the status quo, which are then enshrined in protocols attached to the Treaty. This does not alter the fact that the strong presumption is that the acceding state takes the package that is on offer as constituted by acceptance of the acquis communautaire. This is so notwithstanding the fact there will inevitably be negotiations as to how much the acceding state pays into the EU and how much it gets out. The ‘prix fixe’ nature of accession is further reinforced by the fact that there is in reality no firm line between the agreement on accession and the trade dimension thereafter, since the former embraces the latter: acceptance of EU membership entails acceptance of the EU Treaty rules on the single market, competition, state aids, and the like. The very fact that a new member is joining the club is then reflected in the voting rules. Unanimity in the Council and consent of the EP constitute the twin legitimacy required for conferral of membership status, combined with ratification in accord with the constitutional requirements of each Member State, thereby securing democratic imprimatur at national level.

Withdrawal is more complex and is akin to an ‘à la carte’ deal. This is so in relation to both aspects of withdrawal. The withdrawal agreement has to be crafted afresh on each occasion, there is no boilerplate on which to rely. This is so notwithstanding the fact that if there should be any withdrawal agreement post-Brexit then it is likely that it will cover some of the same terrain as that negotiated between the EU and the UK. There are however likely to be significant differences, such as the absence of the Northern Ireland problem. The ‘à la carte’ nature of withdrawal is further evidenced by the trade dimension. There is no a priori reason why there should be any particular trade deal between the EU and the withdrawing state, nor that there should be any particular deal on future relations more broadly conceived. There is a significant range of possibilities in this regard, which are subject to negotiation and contestation between the two sides.⁶ This is starkly exemplified by Brexit,⁷ wherein the bargaining position of the UK government led by Boris Johnson⁸ shifted markedly from that championed by Theresa May.⁹ The difference between accession and withdrawal plays out yet again in the voting rules. Qualified majority voting, not unanimity, is required for the withdrawal agreement, the rationale being that if a state wishes to leave then it suffices that the precise terms of the departure are accepted by a qualified majority. By way of contrast, unanimity is stipulated for an extension of the negotiations beyond two years, since the implications of doing so may be significant, hence the requirement of consent by all Member States.

III. SUBSTANTIVE DIMENSION: MEMBERSHIP AND MEMBER STATES

A. Member States: Rights and Duties

Membership is undertaken by the acceding state and has consequences for the state qua Member State when its application is accepted. When a Member State accedes to the EU it is subject to the rights and duties embodied in the Treaty. That is axiomatic. We can, however, press further and enhance our understanding of the nature of the rights and duties contained in the constituent Treaties. This is more especially so since they are not conceptually uniform. Consider, by way of example, the duties that inhere in membership. We can distinguish between the following types of commitment.

1. Discrete Primary Duties

The Treaties contain many discrete primary duties, which capture much of the core substance of EU membership. They are exemplified by the obligations associated with the single market, the state aid regime, and the environment. The duties are central to the rationale for the EU, albeit in different ways.

The commitments flowing from the single market include those relating to customs duties, tariffs, discriminatory taxation, and the four freedoms.¹⁰ They embody the economic essence of a common market, whereby resources can be used in an optimal manner throughout the EU, shorn of limitations based on nationality. If, for example, there is unemployment in a particular part of the EU, manifest in an excess supply over the demand for labour, then workers should be able to move freely to other parts of the EU where labour is in relative short supply, thereby benefiting the individual workers and enhancing the overall value of labour within the EU. The four freedoms also express social objectives, in particular relating to free movement of persons, whereby the Treaty provisions are designed to facilitate the social interaction of the EU citizens.

The discrete responsibilities contained in the rules concerning state aid are equally central to the EU. The single market is predicated on the creation of a level playing field, wherein individuals and corporations compete on equal terms. This objective would perforce be undermined if Member States could furnish economic advantages to their own companies, hence the strict regime of control whereby such aid is prima facie prohibited, subject to limited exceptions policed by the Commission.¹¹

The Treaty rules relating to the environment, and legislation enacted thereunder, provide a further example of a primary Treaty duty imposed on Member States. The eclectic rationale for the inclusion of such responsibilities is evident from Article 191 TFEU, which states that EU environmental policy shall contribute to preserving, protecting, and improving the quality of the environment; protecting human health; prudent and rational utilisation of natural resources; and promoting measures at international level to deal with regional or worldwide environmental problems, and in particular combating climate change. There was moreover a particular specific catalyst for according the EU competence over the environment. The effects of pollution do not stop at borders. Lax environmental law in one country can therefore create externalities, the costs of which are borne by those in other states. This in turn means that the costs to industry of weak environmental laws will be lower than for countries where the standards are correspondingly higher, thereby distorting the level playing field that is meant to be at the heart of the EU.

2. General Primary Objectives/Duties

The constituent Treaties also contain what may be termed general primary objectives/duties. How far these translate into obligations that can be imposed directly on a particular Member State will be considered in due course. Articles 2 and 3 TEU are the principal, albeit not sole, source of such objectives/duties, and must be read with related Treaty provisions.

Article 2 TEU provides that the EU is founded on the values of respect for human dignity, freedom, democracy, equality, the rule of law, and respect for human rights, including the rights of persons belonging to minorities. It further provides that these values are common to the Member States in a society in which pluralism, non-discrimination, tolerance, justice, solidarity, and equality between women and men prevail. Article 2 has, as will be seen below, been drawn on in the discourse, academic and judicial, concerning rule of law backsliding by certain Member States.

Article 3 TEU sets out the EU's underlying objectives, which are to be attained ‘commensurate with the competences which are conferred upon it in the Treaties’.¹² These are framed throughout in terms of collective obligations on the EU, as expressed through language such as the EU shall ‘provide’ or ‘establish’. The EU however operates through the instrumentality of its Member States, in the Council and the European Council. An obligation on the EU to ‘provide’, ‘establish’, or ‘promote’ the values mentioned in Article 3 can therefore only be attained through the collective exercise of Member State action in these EU institutions. This collective political obligation exists irrespective of whether legal obligations flow from such provisions. It is moreover important when thinking about legal obligations to disaggregate the direct imposition of legal duties based on Article 3, from the fact that this Treaty article may inform the interpretation of other more discrete Treaty provisions.

Article 3(1) is framed in terms of the Union's aim being to promote peace, its values, and the well-being of its peoples. Article 3(2) speaks to the issues dealt with by the Area of Freedom, Security, and Justice (‘AFSJ’). Article 3(3) states that the EU shall ‘establish’ an internal market based on ‘balanced economic growth and price stability, a highly competitive social market economy, aiming at full employment and social progress, and a high level of protection and improvement of the quality of the environment’.¹³ It further provides that the EU shall ‘combat’ social exclusion and discrimination, and shall ‘promote’ social justice and protection, equality between women and men, solidarity between generations, and protection of the rights of the child. Article 3(3) also attests to the importance of economic, social, and territorial cohesion, and solidarity among Member States, thereby providing the foundation for the Structural Funds, and the importance of cultural and linguistic diversity. Article 3(4) stipulates that the EU shall ‘establish’ an economic and monetary union, while Article 3(5) extols the EU to ‘uphold’ and ‘promote’ its values and interests in its relations with the wider world,¹⁴ which should be read with Article 23 TEU.

3. Discrete Secondary Remedial Powers and Duties

It is common in national legal systems to distinguish between primary duties resulting from a legal relationship, such as a contract, and secondary obligations that flow from breach of the primary obligation. EU law is no different as a matter of principle in this respect, although identification of the remedial consequences of breach of the primary obligation may be more complex than in national legal orders for reasons set out below.

There are a number of discrete Treaty provisions that afford the EU power to impose remedies for breach of primary obligations specified in the constituent Treaties, the following being merely examples. Thus, Article 103 TFEU empowers the making of regulations or directives designed, inter alia, to ensure compliance with the prohibitions in Article 101(1) and in Article 102 TFEU, by making provision for fines and periodic penalty payments. Article 105 TFEU further empowers the Commission to investigate breach of Articles 101–102 TFEU concerned with competition, to decide that an infringement has occurred and to propose appropriate measures to bring it to an end. If the infringement is not terminated, the Commission can authorise Member States to take measures, the conditions being determined by the Commission, to remedy the situation.

The European Central Bank can impose fines or periodic penalty payments on undertakings that fail to comply with obligations under its regulations and decisions.¹⁵ In a different area yet again, the Commission makes the operative decision as to whether state aid is compatible with the functioning of the internal market. Such decisions are legally binding and breach thereof can be enforced through an expedited enforcement action under Article 258 TFEU.¹⁶ Remedial power does not adhere solely to the Commission. The Council is afforded remedial capacity in certain instances, most notably in relation to the excessive deficit procedure.¹⁷

4. General Secondary Remedial Powers and Duties

The constituent Treaties also contain more general remedial powers that impose correlative obligations on Member States. Three such powers and duties can be identified in the Treaty.

The first and most obvious is the remedial system in Articles 258–260 TFEU. It embodies the regime of public enforcement of EU law that has been central to the Treaty from the outset. Enforcement is trusted to the Commission, which has discretion as to which cases it wishes to pursue, given the limited resources that it has to devote to the task. It was a relative novelty when the Rome Treaty was ratified, since it constitutes compulsory adjudication of a kind that was not common in international treaties, more especially because Member States cannot derogate or limit such adjudication. It has been strengthened over time, most importantly through the addition of what is now Article 260 TFEU, which enables the CJEU to impose penalty payments where a Member State has not complied with a prior ruling that it acted in breach of the Treaty.¹⁸

The second general remedial provision is to be found in Article 19(1) TEU, second paragraph, which states that Member States shall provide remedies sufficient to ensure effective legal protection in the fields covered by Union law. The EU Treaty does not contain detailed rules concerning the national remedies to be provided for breach of EU law. This gave rise to the complex body of case law as to the extent to which the CJEU controls national remedial protection. While the foundational precept is that if there are no EU rules specifying the remedies and procedures available for breach of EU law then the matter is left to national law, this has been qualified. There are long-established criteria requiring equivalence between remedial relief in relation to national and EU measures, and that the national rules do not render it impossible to secure relief for violation of an EU right. There are, in addition, requirements to ensure that the remedy is adequate, proportionate, and that it provides effective judicial protection. The requirement of effective judicial protection was reinforced by inclusion in Article 47 of the Charter of Rights. The relationship between the constraints on national procedural autonomy cast in terms of equivalence/effectiveness and that framed in terms of effective judicial protection is itself interesting.¹⁹

Article 19(1) is particularly apposite for the present analysis, since it tells us much about the relationship between the CJEU and the Treaty. Viewed from one perspective, Article 19(1), second paragraph, which was included in the Lisbon Treaty, constitutes a retrospective vindication of the CJEU's case law imposing limits on national procedural autonomy. The CJEU had imposed such limits as a matter of first principle. While the Lisbon Treaty did not prescribe detailed secondary remedial duties on Member States, it did incorporate the general secondary remedial obligation now found in Article 19(1), which legitimated what the Court had done hitherto. Viewed from a related perspective, Article 19(1) then proved, as we shall see below, a fertile foundation on which the CJEU grounded its case law concerning the importance of national judicial independence as one weapon in the armoury against rule of law backsliding by Member States.

The third general remedial duty is to be found in Article 4(3) TEU. The principle of sincere cooperation demands that the EU and Member States shall assist each other in carrying out Treaty tasks. It is incumbent on the Member States to take any appropriate measure, general or particular, to ensure fulfilment of Treaty obligations, or duties resulting from acts of the EU institutions of the Union. The Member States must moreover refrain from any measure that could jeopardise attainment of the EU's objectives.

B. Link between Primary and Secondary Duties

There are perforce multiple examples of the link between primary and secondary duties. The application of a secondary remedial duty presumes, by definition, that there has been a breach of a primary duty. There is then nothing uncommon about the linkage, it is, to the contrary, the norm. This is more especially so when breach of a discrete primary duty leads to the application of a discrete secondary remedial obligation. The connection is, however, more interesting where it is between a general primary duty and a general secondary remedial commitment. There will inevitably be greater judicial creativity in interpreting the demands of the primary duty, and the way in which it is backed up by the secondary remedial provision. The incentive to make the connection is enhanced by the importance of the substantive issue for EU membership more broadly conceived. The discussion that follows considers two powerful examples of this connection.

1. Rule of Law and Article 19 TEU

There has been rule of law ‘backsliding’ by some Member States, as attested to in the large literature on the subject.²⁰ This is especially significant when viewed through the lens of EU membership. It can undermine the application of EU law within that particular Member State, and it can more generally sap the EU's moral authority, insofar as a non-democratic state might be a member of the Union. There was then a strong incentive for the CJEU to forge the link between violation of Article 2 TEU and the remedial provision in Article 19(1) TEU.

The EU's rule of law problem is primarily, although not exclusively, concerned with challenge to the independence of the judiciary in certain Member States, notably Poland and Hungary. An independent judiciary is an especially significant component of the rule of law. Thus, it is central to any conception of the rule of law that the government should act on a basis that is deemed valid by that legal system. If the government or legislature exceeds the boundaries of its lawful authority, then its action will be null or invalid. There must be independent courts to assess, in an objective manner, whether the limits on lawful authority have been exceeded. If the courts lack such independence, or are subservient to the will of the political branch of government, then there is a real danger that fundamental limits on the scope of political power will be ignored. The judiciary must also be independent to give legal effect to other precepts of the rule of law. Thus, independent courts in most legal systems have a range of tools at their disposal to deal with, for example, retrospective laws, those which are unclear, or rules that inhibit access to court. If the courts lack independence, then they will not protect these principles that comprise the rule of law. Whether they fail wholly in this respect depends on the extent to which their independence is compromised.

The centrality of judicial independence to the rule of law is especially pertinent in the EU context. Mutual trust between national courts is integral to the European Arrest Warrant regime that constitutes the core of the AFSJ. In more general terms, national courts are fundamental to the regime of EU adjudication. They have the obligation to conform to, and apply, EU law within their jurisdiction. They are also central to the flow of cases upwards to the CJEU, via preliminary references in Article 267 TFEU. If national courts lack independence then the regime of EU adjudication will suffer in both respects. National courts that lack independence may not apply EU law correctly, and they may restrict the flow of preliminary references where there are challenges to national action that is contrary to EU law.

The EU has deployed a plethora of institutional, administrative, and legal mechanisms to combat the rule of law problem.²¹ This has included use of preliminary rulings and infringement actions. The legal reasoning in such cases is of especial interest in the present context, since it reveals the conjunction of a general primary duty to comply with the rule of law, derived from Article 2 TEU, with a general secondary remedial duty to provide effective legal protection based on Article 19(1) TEU.

In Panicello,²² the CJEU articulated the twin dimensions of judicial independence for a national court making a preliminary reference. The external dimension demanded that the court exercised its functions wholly autonomously, without being subject to any hierarchical constraint or subordinated to any other body, and without taking orders or instructions from any source whatsoever. The internal dimension required impartiality, objectivity, and the absence of any interest in the outcome of the proceedings apart from the strict application of the rule of law, thereby ensuring a level playing field for the parties to the proceedings.

In subsequent cases, the CJEU built on this external and internal conception of judicial independence and forged the link with Article 19 TEU. This is evident in the Portuguese Judges case.²³ It involved a challenge to a Portuguese law to reduce the budget deficit, which entailed reducing the salary of Portuguese judges. The CJEU concluded that this did not infringe the requirement of judicial independence. However, it held that Article 19 TEU gave ‘concrete expression to the value of the rule of law stated in Article 2 TEU’,²⁴ and that the responsibility for ensuring judicial review in the EU legal order was shared between EU courts and national courts. It followed from the requirement of effective judicial protection that judges in national courts should be independent. This was inherent in the very idea of adjudication and was required to ensure that the preliminary ruling system could operate as intended.²⁵

The link between Article 2 TEU and Article 19 TEU was once again at the forefront of the analysis in Commission v Poland.²⁶ The case involved a challenge to a Polish law that changed the terms on which judges of the Supreme Court worked, and their retirement age, vesting the President of the Republic with wide powers in this regard. The CJEU reiterated the reasoning from the Portuguese Judges case. It held that judicial independence was inherent in the task of adjudication, that it was central to the right to effective judicial protection and the fundamental right to a fair trial. This was of cardinal importance as a guarantee that all EU rights would be protected and ‘that the values common to the Member States set out in Article 2 TEU, in particular the value of the rule of law’, would be safeguarded.²⁷ The requirement of independence meant that the rules governing the disciplinary regime and dismissal of judges should not be used ‘as a system of political control of the content of judicial decisions’.²⁸ Moreover, in Łowicz, while ruling that the preliminary reference was inadmissible because the case did not concern EU law, the CJEU nonetheless made clear that provisions of national law that exposed national judges to disciplinary proceedings if they submitted a preliminary reference to the Court infringed judicial independence and were contrary to EU law.²⁹

2. State Obligations and Article 4(3) TEU

The connection between primary duties and a general remedial obligation is also evident in the jurisprudence concerning Article 4(3) TEU. The connection with conceptions of membership is readily apparent, compelling, and foundational. The principle underlying this case law is that the obligation of sincere cooperation imposes a duty to remedy the breach of the primary obligation. The precise form of the remedy will differ depending on the circumstances. Three brief examples can be proffered here as to the reach of Article 4(3).

First, there is the Francovich jurisprudence.³⁰ The European Court of Justice (‘ECJ’) held that the Treaty created its own legal system from which individuals derived rights as well as obligations. This included rights that flowed from directives, as well as from Treaty provisions. The full effectiveness of such ‘Community rules would be impaired and the protection of the rights which they grant would be weakened’³¹ if individuals could not obtain compensation when the Member State was responsible for breach of such rights. The general primary duty to comply with Community law generated a secondary remedial duty to pay compensation when the rights of individuals were affected. The connection was forged through recourse to effectiveness. The remedial obligation was then reinforced through what is now Article 4(3) TEU, ‘under which the Member States are required to take all appropriate measures, whether general or particular, to ensure fulfilment of their obligations under Community law’,³² which included the obligation to nullify the unlawful consequences of a breach of Community law.

It is moreover clear that Article 4(3) can be used to ground a damages action in a claim brought by the Commission, rather than an individual. Thus, in Commission v UK ³³ the CJEU held that pursuant to the principle of sincere cooperation, the Member States were required to nullify the unlawful consequences of an infringement of EU law. It was for the Member States to take all measures necessary to remedy an infringement of EU law, which included an obligation to compensate for a wrongful act by the Member State that led to loss of the EU's own resources. This reasoning was reiterated in Commission v Netherlands,³⁴ where the CJEU rejected the argument that an obligation to compensate cannot exist if there is no express provision for it in EU law. It held that the obligation to compensate for the loss of the EU's own resources resulting from a wrongful Member State act was ‘merely a particular expression of the obligation, arising from the principle of sincere cooperation, under which the Member States are required to take all necessary measures to remedy an infringement of EU law and to nullify the unlawful consequences of it’.³⁵

Secondly, Article 4(3) is regularly used as the foundation for Member State action to remedy the breach of EU law, irrespective of whether this entails a monetary claim. Thus, the CJEU has repeatedly held that the principle of sincere cooperation in Article 4(3) TEU requires Member States to nullify the unlawful consequences of the breach of EU law. National authorities must therefore take all measures necessary, within the sphere of their competence, to remedy, for example, a failure to carry out an environmental impact assessment by revoking or suspending consent already granted in order to carry out such an assessment.³⁶

Thirdly, Article 4(3) TEU has also been used to limit Member State remedial discretion. This is exemplified by OPR-Finance ³⁷ where the CJEU held that in accord with the principle of sincere cooperation in Article 4(3) TEU that, while the choice of penalties remains within their discretion, Member States must ensure that infringements of EU law are penalised under analogous conditions to infringements of national law of a similar nature and which, make the penalty effective, proportionate, and dissuasive.³⁸

IV. SUBSTANTIVE DIMENSION: MEMBERSHIP AND INDIVIDUALS

The substantive discussion of EU membership has thus far been concerned with the rights and duties that attach to Member States. It is, however, equally important to consider membership from the perspective of the individual. It is axiomatic that it is the state that accedes to the EU. Successful accession nonetheless has important consequences for individuals that speak directly to the conception of membership that underpins the Union. This is readily apparent in relation to direct effect and citizenship, which will be considered in turn.

A. Direct Effect, Rights, and Membership: The Legal and the Political

The seminal decision in Van Gend en Loos was quintessentially about rival conceptions of membership.³⁹ It was not pre-ordained. There is nothing a priori in the concept of direct effect and individual rights. It would be perfectly possible to have a Community wherein the Member States are the repository of rights and duties. Membership conceptualised in these terms is the norm under international treaties, and this vision informed Member State argument before the Court by the Netherlands and the intervening states.⁴⁰ From that perspective, the international treaty norm concerning membership was equally applicable to the newly created European Economic Community (‘EEC’). This was more especially so given that the Treaty contained an enforcement mechanism in Article 169 EEC, which embodied a regime of compulsory adjudication based on public enforcement via the Commission. There was then, on this view, no warrant for departing from traditional conceptions of membership, cast in terms of sovereign Member States being the sole bearers of rights and duties.

The ECJ rejected this vision of membership and in doing so laid the juridical foundations for direct effect. The Court was fully cognizant of what was at stake. Membership is at the forefront of its argument, frames the subsequent analysis, and informs the conclusion. The ECJ held that the EEC Treaty established a Common Market, the functioning of which was of direct concern to interested parties in the Community, which implied ‘that this Treaty is more than an agreement which merely creates mutual obligations between the contracting states’.⁴¹

The Court drew on further argumentation to support this conception of membership.⁴² It called in aid the Treaty preamble, which referred not just to governments but to peoples; it noted that the exercise of Community powers could affect individuals and not just Member States; it pointed to the fact that individuals participated in Community governance through the European Parliament and the Economic and Social Committee; and it used Article 177 EEC to show that ‘the states have acknowledged that community law has an authority which can be invoked by their nationals before those courts and tribunals’.⁴³

The Court's conception of membership, afforced by the preceding arguments, then informed the famous conclusion. The Community had, said the Court, constituted a new legal order of international law for the benefit of which the states have limited their sovereign rights, ‘and the subjects of which comprise not only Member States but also their nationals’.⁴⁴ It followed that Community law not only imposed obligations on individuals, but gave them rights, which could arise either from express Treaty provision, or from Treaty obligations imposed on individuals, Member States, or Community institutions.

The Court then exemplified this through analysis of Article 12 EEC, the Treaty provision at stake in the instant case. It contained a clear and unconditional prohibition, imposed a negative not a positive obligation, and did not require any further action to be taken at national level. The very nature of this prohibition made ‘it ideally adapted to produce direct effects in the legal relationship between Member States and their subjects’.⁴⁵

It remained only for the ECJ to dismiss the argument that direct effect was not required because of the remedial mechanism in Articles 169–170 EEC. The ECJ deployed a classic mode of legal reasoning used by courts the world over: the fact that there is one kind of remedial mechanism does not preclude the existence of an alternative means of redress. Thus, the existence of Articles 169–170 if a Member State infringed its Treaty obligations did not ‘mean that individuals cannot plead these obligations, should the occasion arise, before a national court’.⁴⁶ This was more especially so because reliance solely on Article 169 and 170 ‘would remove all direct legal protection of the individual rights of their nationals’.⁴⁷

The rest, as they say, is history. Van Gend en Loos laid the foundation for direct effect, which has been developed in many subsequent cases, coupled with a vibrant academic literature cataloguing and commenting on all the twists and turns of the jurisprudence. This is not the place for detailed exegesis on the case law and commentary thereon. Suffice it to say the following, concerning the importance of the decision in relation to conceptions of EU membership and the remedial dimension of EU law.

Van Gend en Loos was of seminal importance for the conception of EU membership that it articulated, and we lose sight of this at our peril. There is room for all manner of debate concerning, for example, the scope of direct effect, its impact on national courts, constitutionalisation of EU law, and whether the grant of individual rights can on occasion have negative effects for the development of the EU. Discourse on such matters should not, however, mask the zero-sum choice faced by the Court in 1962: membership could be conceptualised in traditional international terms, wherein the sole bearers of rights and duties were the Member States, or it could be conceived more broadly, so that individuals could also have rights as well as duties. The Court opted for the latter and the impact was profound. It gave EEC law immediate relevance for individuals, who could derive rights from Community law and enforce them via their national courts. The EEC Treaty was not the legal preserve solely of states, as was so for most international treaties. Individuals were legal subjects in their own capacity and made extensive use of direct effect as attested to by the case law. Normative supranationalism, fuelled through direct effect, helped to prevent the stagnation of Community law during the period of Community malaise, when decisional supranationalism, through the Council, was difficult.⁴⁸

Van Gend en Loos was also of seminal importance for membership in remedial terms. The EEC Treaty was predicated on public enforcement. It was sophisticated in this respect, going beyond the enforcement mechanism found in most international treaties. There are, however, very real limits to public enforcement. The Commission only has limited capacity and cannot in reality prosecute more than a fraction of infringements, a problem that has grown over time with the increase in the number of Member States and the volume of Community legislation. There are structural limits to public enforcement, since it can, in accord with the wording of Article 169, only be used against a Member State, thereby leaving untouched those instances where a private party violated EU law. The remedy pursuant to Article 169 was weak, since before the Maastricht Treaty, it was merely declaratory. Direct effect is a mechanism of private enforcement, in the sense that it lies in the hands of the affected individual. It alleviated or cured the limitations on public enforcement. Direct effect created a very large body of potential enforcers of EU law, equivalent to all those who benefited from a particular EU right, thereby alleviating the problem that the Commission only had limited enforcement resources. Direct effect could operate horizontally, as well as vertically, thereby alleviating the structural limit to public enforcement under Article 169. Direct effect was more potent in remedial terms, since the action began and ended in the national court, which could give the claimant the relief sought, such as restitution of the illegal import duties in Van Gend. Private enforcement through direct effect thereby complemented public enforcement through Article 169, as recognised by the ECJ in Van Gend, which concluded that ‘the vigilance of individuals concerned to protect their rights amounts to an effective supervision in addition to the supervision entrusted by Articles 169 and 170 to the diligence of the Commission and of the Member States’.⁴⁹

B. Citizenship, Rights, and Membership: The Political and the Legal

The Treaty provisions on citizenship speak to conceptions of EU membership. There is, as with direct effect, no a priori reason for such provisions. It is perfectly possible to have a Community with Member States thereof, without Arts 20–25 TFEU on EU citizenship. This is readily apparent given that they were only introduced in the Maastricht Treaty.⁵⁰ The change was not fortuitous, being driven by the desire to thicken the political and civic bond created by the EU.⁵¹ It resonated with ideas concerning broader European political integration that dated from the post-war period. It fostered the idea that citizenship did not have to be tied exclusively to the nation state. It connected with the desire for greater democratisation within the Community, since the identification of Member State citizens as Community citizens thereby spoke to the fact that the EC was not solely about trade.

The more particular ramifications of citizenship, and the conception of membership thereby conveyed, was the result of an admixture of Treaty provision, CJEU rulings, and EU legislation. The complex interplay between these modalities shaped the contours of EU citizenship over the subsequent three decades. There was nothing preordained about the outcome, nor was the path of development linear. To the contrary, there were many twists and turns en route. It is impossible within the confines of this article to relay this detailed story. The object is rather to convey the symbiotic interaction between Treaty provision, CJEU rulings and EU legislation concerning citizenship, and the idea of membership embodied therein.

The Treaty provisions were expressive of Member State choice as to the content with which to imbue the nascent idea of Community citizenship. It is captured in Article 20 TFEU and elaborated more fully in Articles 21–24 TFEU. It is clear from Article 20(1) that EU citizenship is parasitic on citizenship of a Member State. The more particular citizenship rights are elaborated in Article 20(2): the right to move and reside freely within the territory of the Member States; the right to vote and to stand in EP elections and municipal elections in their Member State of residence, under the same conditions as nationals; rights to diplomatic protection; and rights concerning use of any Treaty language when addressing EU institutions and other bodies. While the Member States were willing to amend the Treaties to include citizenship, they were cautious, as attested to by the limited rights listed in Article 20. This caution was reinforced by the concluding sentence in Article 20, which states that the rights ‘shall be exercised in accordance with the conditions and limits defined by the Treaties and by the measures adopted thereunder’. The Treaty amendment nonetheless impacted on notions of Union membership, in particular through the right to move and reside freely within the territory of the Member States. The introduction of an autonomous Treaty right to move and reside within any Member State, even if the person was economically inactive, meant that the consequences of membership for EU citizens had a broader impact than hitherto, with an increased emphasis on the civic and social interaction of the peoples of Europe.

The significance of this Treaty amendment was dependent in part on the ECJ's reaction. A restrictive interpretation would crab and confine citizenship, a more expansive reading would imbue it with greater force. The ECJ signalled the latter approach in Grzelczyk, where it stated that ‘Union citizenship is destined to be the fundamental status of nationals of the Member States’.⁵² The more precise meaning of this oft-repeated Delphic phrase is contestable, and has shifted over time. This should not, however, cause us to forget the judicial choice in thus elevating citizenship, and reiterating it in subsequent jurisprudence. While rights to move and reside in certain instances had previously been enshrined in Community legislation, the shift to an autonomous Treaty right had juridical consequence.

This was exemplified by Baumbast,⁵³ where the ECJ held that Article 20(1) TFEU conferred a directly effective right on EU citizens to reside in a host Member State, regardless of whether they are employed or self-employed. The fact that Article 20(1) TFEU had moved the rights of such citizens from a legislative to a Treaty footing was not merely a symbolic change. It had legal consequence, since the ‘limitations and conditions’ accepted by the Treaty on the rights of movement and residence had to be applied in a proportionate way and could not undermine the right of residence conferred directly by the Treaty. This was exemplified again in the case law on internal situations, where the CJEU in cases such as Kohll ⁵⁴ and Ruiz Zambrano ⁵⁵ interpreted the precept that EU law does not apply to wholly internal situations narrowly in the light of the citizenship right to move and reside in Article 21 TFEU. The juridical consequences flowing from the fact that citizenship was accorded a Treaty foundation are further evident in relation to citizen access to social benefits in cases such as Martinez Sala ⁵⁶ and Trojani.⁵⁷

However, the complex jurisprudence also reveals contestation as to the legal consequences of citizenship, and the picture of membership that it reveals. Thus, for example, the scope of Ruiz Zambrano has been scaled back in cases such as Dereci ⁵⁸ and McCarthy,⁵⁹ and the extent to which citizenship can generate claims to social benefits was qualified in cases such as Dano ⁶⁰ and Alimanovic.⁶¹ Judicial equivocation in this regard is unsurprising. It reflects at one stage removed uncertainty as to the degree of solidarity that should attach to EU citizenship.⁶² This is a recurring feature in other parts of EU law. The constituent Treaties are in certain respects akin to incomplete contracts, and this is so notwithstanding their combined length. There are perforce issues that cannot be fully elaborated in a Treaty provision, which must then be fleshed out through legislation and judicial interpretation. The devil is always in the detail, which will be worked and re-worked through legislation and adjudication. The resulting content of citizenship will represent the conception of membership that the EU is willing to accept at any point in time.

V. SUBSTANTIVE DIMENSION: MEMBERSHIP AND DIFFERENTIATED INTEGRATION

A. Membership: Factors Driving Differentiated Integration

The introduction of provisions concerning enhanced cooperation in the Treaty of Amsterdam was the catalyst for early academic discourse on differentiated integration.⁶³ The particular Treaty provisions have not proven fertile ground for practical implementation of differentiated integration and have only been used on few occasions. There has nonetheless been considerable discussion of differentiated integration more broadly, in relation to, for example, the euro area, Schengen, defence cooperation, or differential application of Treaty provisions when operationalised through EU legislation. This is attested to by an extensive academic literature from political science and law.⁶⁴

Rival definitions abound. There is, however much, to be said for the distinction between ‘multiple speeds’, ‘federal core Europe’, and ‘flexibility à la carte’, articulated by Thym.⁶⁵ Thus, the key characteristic of the ‘multiple speeds’ model is, as the nomenclature suggests, that all Member States remain bound by the final objective, but the time frame for realising this differs as between Member States. Viewed from this perspective, ‘multiple speeds’ is a transitional phenomenon, developed to cope with distinctions between Member States, primarily, albeit not exclusively, in terms of the stage of their economic development. The ‘federal core Europe’ model is predicated on continuing closer integration, with the consequence that the willing Member States press ahead through whatever modality best effectuates this end, whether this be enhanced cooperation, Treaty reform, or development outside the strict letter of the constituent Treaties. By way of contrast, the ‘flexibility à la carte’ model rejects the federal vision, placing emphasis on Member State freedom to decide on their desired degree of participation.⁶⁶

The present discussion is concerned with the relationship between membership and differentiated integration. It is helpful to clear the ground in this respect by focusing on first principles. There is no a priori reason why the concept of membership must entail uniformity across all domains of EU law. There is nothing untoward in an organisation having terms of membership that differ in certain respects, with different obligations attaching thereto. There is, however, equally no a priori reason why the concept of membership must involve differentiated integration. Many organisations function with flat rights and obligations that pertain to all.

The factors that circumscribe uniformity of application and drive differentiated integration are readily identifiable. Thus, other things being equal, the greater the substantive reach of the Treaty regime, and the greater the number and heterogeneity of Member States, the greater incentive for some differentiated integration. The more demanding the EU obligations, and the more intrusive the impact on national sovereignty, the greater the incentive and pressure for some special treatment by some Member States, whether this be in relation to, for example, EMU or AFSJ.

B. Membership: Factors Circumscribing Differentiated Integration

Differentiated integration is regarded as a distinct topic in its own right. The relationship with membership is perceived to be secondary, such that differentiated integration defines the more particular conception of membership that Member States are willing to embrace. There is some force in this picture. It is however incomplete, insofar as it conceives of membership solely in terms of the output of differentiated integration at any point in time. The discussion in the preceding sections of this article reveal the significance of conceptions of membership over and beyond this. There is, moreover, value in viewing differentiated integration through the lens of membership, since it is only by doing so that we can properly understand the factors that circumscribe such differentiation.

1. Decisional Constraints and Formal Legitimacy

Membership properly understood imposes decisional constraints on differentiated integration. It demands that the uneven application of Treaty obligations as between the Member States has foundation in a source that is regarded as formally legitimate within that legal order. This is so irrespective of the content of the differential application or reach of EU law.

This source may be found in opt-outs included in the constituent Treaties, exemplified by those that the UK had in relation to the social chapter and the euro. It may be located in the wording of Treaty provisions, which afford Member States latitude as to whether to sign up to certain aspects of EU policy, an example of which can be found in Articles 42(6) and 46 TEU, dealing with permanent structured cooperation in the common defence policy. The formal legitimacy may be forthcoming from EU legislation, where the differential application is the result of choice made pursuant to the procedure for making such legislation, provided that the differentiation is not incompatible with the empowering Treaty provisions pursuant to which such legislation is made.

By way of contrast, formal legitimacy is lacking where the differential application is the result of administrative practice, power politics or inadequate enforcement. We do not live in a perfect world. There will, therefore, always be instances where this occurs. This can be acknowledged, but it does not thereby alter the point made here, nor does it cure the infirmity that flows from the absence of formal legitimacy.

2. Functional Constraints and Substantive Effectiveness

Membership imposes functional limits on differentiated integration. The Member States join a club when acceding to the EU. Realisation of the benefits of membership may be dependent on a commonality of application of the relevant provisions, which is undermined through their differential application. This may be so irrespective of whether there is formal legitimacy for the differentiation. There can, therefore, be negative externalities flowing from differential integration.

This can occur where the differential application of EU law operates vertically, as the result of discretion given to Member States as to how to discharge EU legislation. Thus, for example, the EU might decide to pursue a subsidiarity strategy. It might do so by leaving certain aspects of the regulatory regime to be dealt with at the national level; by specifying EU rules to govern all aspects of the regulatory schema, but doing so at a relatively high level of generality, thereby leaving more scope for national input and variation; or by an admixture of both strategies. Lighter touch intervention through directives rather than regulations has been designed to foster subsidiarity. The reality is that in many areas, such as telecommunications, energy, agriculture, the structural funds, and financial services regulation, the desire to foster subsidiarity, either by leaving certain aspects of the regulatory regime to national rules, or through EU rules that govern the issues but are set at a high level of generality so as to allow for national choice, have led to regulatory failure, with the consequence that the rules have had to be revised and the level of EU control ratcheted-up.⁶⁷

There can also be functional problems when differential integration operates horizontally. If differential treatment is allowed in one Member State it may have negative spill-over effects on the application of the regulatory provisions in other Member States, which is somewhat paradoxical, given that part of the initial rationale for creation of the EU was to deal with externalities created by state regulatory activity. There can, in addition, be more general functional problems, flowing from overlap between regulatory spheres. The real world does not divide neatly into hermetically sealed subject matter areas. Thus, there have, for example, been concerns that regulation of the euro area can impact on non-euro Member States. This concern has been recognised by the European Council, which emphasised that the process towards deeper economic and monetary union should be characterised by ‘transparency towards Member States which do not use the single currency and by respect for the integrity of the Single Market’.⁶⁸

3. Normative Constraints and Substantive Legitimacy

Membership also imposes normative constraints on differentiated integration. The nature of such constraints will perforce be affected by the rationale for such distinctive treatment. Thus, the ‘flexibility à la carte’ model would draw the normative constraints differently from the other two models since the difference in treatment would be reflective of Member State choice as to their desired degree of participation.

There are nonetheless normative constraints flowing from membership that pertain, irrespective of the model of differentiated integration espoused. This is especially so in relation to the common core of EU values. The detail of this common core may be contestable, but that does not undermine the point being made here. Differential integration cannot legitimate infringement of the values in Article 2 TEU, and it cannot be used to justify rule of law backsliding.⁶⁹

Nor can it be used to undermine the key features of the single market and the level playing field that it embodies. Member States share the benefits and burdens of membership. Differential integration can generate free rider problems, whereby Member States enjoy the benefits, without the corresponding obligations. This is in part the rationale for the supremacy of EU law, since if Member States could disavow EU regulatory provisions that they disliked, it would thereby render EU law unequal in its application.

C. Membership: Differentiated Integration, Structure, and Effect

The rich discourse on differentiated integration is concerned not only with the unpacking of the status quo, but also with possible ways forward. There are multiple suggestions in this regard.

Thus, for example, Piris distinguished sharply between a two-speed EU, and other versions of differentiated integration, whether cast in terms of variable geometry or multiple speeds. He proposed a two-speed Europe predicated on the same group of Member States acting as the EU vanguard, which accepted the same hard-core of EU obligations across different subject-matter areas:⁷⁰ ‘the primary principle of an “avant-garde” group would be that all participating states should be fully committed to participate in all areas of cooperation, no areas being optional’.⁷¹ The principal subject matter areas were: the economic component of EMU; security and defence policy; aspects of justice and home affairs; social policy; taxation; environmental protection; public health; culture and education; and certain procedural aspects of foreign policy implementation.⁷² This would be effectuated through a new Treaty additional to the Lisbon Treaty, coupled with a novel institutional structure that replicated the roles played by the existing EU institutions.

There have also been very different proposals predicated on a multi-speed EU, such as that advanced by Schmidt,⁷³ where the focus is on a soft-core Europe, made up of the multiple clusters of Member States that ‘overlap in their participation in the EU's many policy communities’.⁷⁴ Commitment to the single market would remain central, and most Member States would belong to a plurality of policy communities, including the Eurozone, security and defence, and Schengen. On this view, a single set of institutions is a key requirement, albeit the decisional rules would require modification for such differentiated integration to operate effectively and legitimately. This vision coheres more closely with a suggestion advanced in the Commission White Paper on the Future of Europe, which saw future development in terms of ‘coalitions of the willing’, whereby some Member States would be at the forefront of cooperation in areas such as defence and security, justice, taxation, and social policy, with the option for other Member States to join when ready or willing.⁷⁵

The desire to introduce greater order into the existing regime is readily explicable. It is therefore unsurprising that there should be efforts to regularise the status quo. It should nonetheless be recognised that the proposals differ not only in detail, but also in relation to the conception of membership that lies therein. Thus, the vision of a two-speed, hard-core Europe is premised on differentiated integration cast in terms of a ‘federal core Europe’, whereas plans for a multiple-speed, soft-core Europe are predicated in part on a ‘flexibility à la carte’ model.

Competing conceptions of membership also underpin some of the difficulties in achieving the desired regularity. Thus, there are conceptual challenges posed by Piris's suggestions for a two-speed Europe, insofar as it is based on the assumption that the same group of Member States accept the same obligations across the different subject matter areas that are covered. The reality is that Member State preferences and interests in relation to each of these areas may well diverge. The vision of a two-speed Europe with all Member States in the vanguard agreeing on all obligations that form part of the new schema is likely to fracture into a multi-speed system that is not very different from that which currently exists. There are, moreover, very considerable difficulties, practical and conceptual, with suggestions for a separate Treaty, and distinctive institutional structure.⁷⁶

Competing conceptions of membership also underpin some aspects of the Schmidt model, notwithstanding that the shift from the status quo is less dramatic. We should not, for example, underestimate the problems attendant on suggestions for variation in voting rules depending on subject matter, and even more so in the related recommendations for according votes to non-Member States on certain issues.⁷⁷ The proposal⁷⁸ for allowing a Member State to exercise a red card in relation to new EU legislation that is rejected by the national Parliament is also highly problematic, notwithstanding the fact that it is qualified by the caveat that this would not be allowed if the opt-out would threaten the viability of the policy, or give an untoward advantage to that state. This would provide fertile ground for endless contestation as to whether the caveat to the opt-out was triggered. It is, moreover, grounded on assumptions about membership that are controversial. It is one thing to suggest, in accord with ‘flexibility à la carte’, that a Member State can exercise choice as to whether to accept, for example, advanced defence integration. It is quite another to interpret this model such that a Member State has a prima facie opt-out from new EU legislation duly enacted, because the national Parliament dislikes it.

There is, moreover, a paradox underpinning suggestions for reform. The status quo is messy, but relatively stable. The substantive content of EU Membership can vary depending, inter alia, on the degree of differentiated integration. The precise content of EU membership in any single Member State, and as between Member States, is dynamic, varying across time. Common institutional and decisional rules apply, subject to any ad hoc amendment thereof demanded by differentiated integration. The suggested reform is designed to be neater, but may be less stable. There are very real difficulties, conceptual and pragmatic, in attempts to formalise differentiated integration through concentric circles, or two-speed Europe. These difficulties are exacerbated if the institutional structure varies depending on the degree of integration, and this is also true, albeit to a lesser extent, when there are complex voting rules that vary depending on the nature of the subject matter, more especially because EU legislative initiatives do not necessarily fit into neat, pre-ordained categories.

VI. SUBSTANTIVE DIMENSION: WITHDRAWAL

The preceding discussion has focused on the substantive dimension of membership. The focus now shifts to withdrawal.

A. Article 50: Voluntarism and State Sovereignty

Article 50(1) is premised on a Member State deciding to withdraw in accord with its own constitutional requirements. The emphasis at the outset of the Article 50 process is therefore firmly on voluntarism and state sovereignty: it is the Member State in exercise of its sovereign will that applied to join the EU, and it is the Member State in pursuance of that same will that decides to leave, in accord with the constitutional requirements pertaining in that Member State.

It was contestation in this regard that gave rise to litigation in the UK in the Miller case.⁷⁹ The salient issue was whether the constitutional requirements in the UK allowed the Prime Minister to give notice to withdraw from the EU pursuant to the prerogative power relating to the making of treaties. Prerogative power is a species of non-statutory discretionary power that exists in certain areas, the exercise of which resides with the executive. The Prime Minister contended that this enabled her to give notice to withdraw without recourse to Parliament and without the need for statutory authorisation. This argument was rejected by the Supreme Court, which held that prerogative power was subject to limits that were applicable in the instant case, such that notice of withdrawal could only be given with statutory authority.

The Miller litigation in the UK was prompted, in part, by concern that the notice of withdrawal might be irrevocable, in the sense that when it was given it could not be unilaterally revoked by the withdrawing state, but only with the agreement of the other Member States. There was a lively academic debate on this issue, which was resolved in the Wightman case,⁸⁰ where the CJEU decided that the notice of withdrawal could be revoked unilaterally. The judgment was informed by conceptions of membership and ceasing to be a member that were grounded in ideas of voluntarism and state sovereignty.

Thus, the CJEU held that Article 50 TEU provided that any Member State could decide to withdraw from the European Union in accordance with its own constitutional requirements, from which it followed that the Member State was not required to take the decision in concert with the other Member States, or EU institutions. The decision to withdraw was for that Member State alone ‘and therefore depends solely on its sovereign choice’.⁸¹ The sovereign nature of the right of withdrawal enshrined in Article 50(1) TEU supported the conclusion that the Member State could revoke the notification of its intention to withdraw from the EU prior to the entry into force of the withdrawal agreement, or if no such agreement was concluded then prior to the expiry of the two-year period in Article 50(3) TEU, subject to any extension thereof.⁸² The CJEU reinforced this conclusion by analogising between accession and withdrawal. Article 50 TEU on withdrawal was, said the CJEU, the counterpart to Article 49 TEU on accession. They were both predicated on the normative assumption that the EU is composed of States that voluntarily committed to the values in Article 2 TEU,⁸³ from which the CJEU then drew the following conclusion: ‘given that a State cannot be forced to accede to the European Union against its will, neither can it be forced to withdraw from the European Union against its will’.⁸⁴

B. Article 50: Bilateralism and Bifurcation

Article 50(1) and the first sentence of Article 50(2) TEU are based on voluntarism and state sovereignty. The remainder of Article 50 is, however, premised on bilateralism and bifurcation. Bilateralism connotes the simple idea that when notice has been given, negotiation concerning withdrawal will be bilateral, conducted between the withdrawing state and the EU. The European Council is in the driving seat for the EU and sets the guidelines for the negotiations from the EU perspective. The Commission, however, produces the detailed draft of the negotiating guidelines for approval by the European Council. The agreement is negotiated in accord with Article 218(3) TFEU and then concluded by the Council on behalf of the EU, with the consent of the European Parliament. Bifurcation, as noted in the earlier discussion, connotes the idea that conclusion of a withdrawal agreement is separate from conclusion of an agreement on future relations between the UK and the EU. Thus, Article 50(2) is framed in terms of the conclusion of a withdrawal agreement, ‘taking account of the framework for its future relationship with the Union’.

There is, in substantive terms, no a priori content to the withdrawal agreement, or the agreement concerning future relations. They are both the result of deliberation between the withdrawing state and the EU. We have, at present, an empirical sample of one, which is, by definition, not a great deal to go on. The Brexit sample is nonetheless indicative of the subjects that would have to be resolved if any other state chooses to exit from the EU.⁸⁵ Thus, the issues of people, finance, transition, and borders will require resolution in any instance of withdrawal.

There is a good deal more room for substantive manoeuvre in relation to the agreement on future relations. The outcome will perforce be dependent on the respective preferences of the withdrawing state and the EU. There is a spectrum of possible outcomes. There might be no-deal, such that even though there is a withdrawal agreement, the withdrawing state and the EU fail to agree any trade deal by the end of the transitional period, with the consequence that the trading relationship is governed by the rules of the World Trade Organization. There might be a hard exit, which signifies that the withdrawing state is not a member of the customs union or the single market, its regulatory standards diverge significantly from those of the EU, and there would be no legal commitment to the maintenance of a level playing field between the two sides. There might be a less hard exit, whereby the withdrawing state is not a member of the customs union or single market, but consciously seeks to maintain regulatory equivalence between its rules and those of the EU, and accepts that there should be a level playing field on matters such as social rights, environmental standards, state aid, and the like. There may also be softer forms of exit, building on analogies with, or membership of, the European Economic Area (‘EEA’), which entails compliance with EU customs and single market rules. The content of the detailed agreement on trade and related matters will be prefigured by the framework for future relations mentioned in Article 50(2) TEU,⁸⁶ although how far this binds the subsequent discourse can be contentious, as evident in the Brexit trade deliberations.

VII. CONCLUSION: REFLECTIONS ON MEMBERSHIP

There is, as is evident from the preceding analysis, much that can be debated concerning particular aspects of membership, whether concerning Member States or individuals. There are, however, three more general issues that warrant elaboration by way of conclusion. They are related but distinct.

A. The Temporal Dimension

There is clearly a temporal dimension to membership, self-evidently so in relation to the number of Member States that constitute the EU, but this would not warrant discussion when reflecting more broadly on the substantive aspects of membership. The focus here is on the temporal dimension, insofar as it affects the membership obligations incumbent on any particular Member State. These may vary for a variety of reasons.

If the particular Member State has an opt-out, it may decide over time to end its privilege, and fold back into the normal Treaty rules that govern that subject matter. It may alternatively decide that it wishes to opt-in to certain EU measures that govern the particular area, assuming that the terms of the opt-out so allow. A particular Member State may elect to sign up to an aspect of EU policy that is not compulsory, such as the permanent structured cooperation in defence, assuming that such a facility has been created. A Member State may choose to join an initiative for enhanced cooperation, either at the outset, or at some time thereafter.

The demands that flow from membership can also alter because of legislative, delegated, and implementing acts made pursuant to the primary Treaty provisions. The EU is not unique among international organisations in having the capacity to make rules in furtherance of the Treaty articles. Its rule-making capacity is, nonetheless, greater than that prevailing elsewhere. This is so in relation to the number, range, and depth of the regulatory provisions, the enactment of which fleshes out the enabling Treaty article, imbuing it with a specificity that it lacked hitherto. The detailed membership rights and obligations increase commensurately, reinforced through doctrines of direct effect and supremacy. The EU legislation may also embody an element of differentiated integration, thereby providing nuance concerning the particular membership duties incumbent on a particular Member State.

B. The Incomplete Agreement

The constituent Treaties are not brief. The TEU has 55 articles, the TFEU has 358, and then there is the Charter of Rights, and the numerous Protocols and Declarations attached to the Treaties. In certain respects, they are, nonetheless, akin to an incomplete agreement, which has implications for the membership rights and obligations that flow therefrom. To some extent this is the reverse side of the temporal coin adumbrated above. Thus, pretty much all Treaty articles contain provisions empowering the making of further legislation, which emerges over time, thereby concretising the more abstract obligations that Member States accepted when acceding to the constituent Treaties.

However, the incompleteness of the Treaty compact goes beyond this in certain areas. There can be contestation as to what the Treaties presently require, which has direct implications for what more concrete measures should be made pursuant thereto, and how the EU courts should interpret the relevant provisions. The contestation may transcend this, in the sense that while it may be accepted that the Treaty does not require a particular solution to a problem, it could nonetheless be interpreted in this manner if the EU institutional players chose to do so. There may, moreover, be disagreement as to whether the Treaty provisions can be read in a certain manner to attain the desired conclusion, and whether the proposed initiative is the best response to the policy problem.

These issues are manifest in diverse areas. They underpin much of the discussion concerning what solidarity demands in the EU. Consider in this respect the debates concerning the financial crisis and the immigration crisis. The divergent voices reflect different assumptions as to what degree of solidarity should be forthcoming to deal with these crises. The salient Treaty provisions may give no certain answer, and there may be dispute as to whether they demand, allow, or prohibit the chosen measure. It is just such discourse that characterises discussion on matters such as Eurobonds/fiscal union, and national quotas for asylum seekers. Similar issues surround discourse concerning citizenship, and the extent to which EU citizens are eligible for social benefits in the host state.

C. The Ascription of Responsibility

Membership is an admixture of rights, duties, powers, and privileges. We ascribe responsibility and blame when duties are broken. This is as it should be. We should nonetheless be mindful of the multiplicity of meaning when we speak of blame, failure, and responsibility. Failure to do so leads to the elision of situations that are in reality distinct.

The paradigm case of blame is indeed straightforward, captured by instances where a Member State fails to comply with a discrete primary Treaty obligation. The resolution of such cases through public enforcement via the Commission, or through private enforcement via direct effect is routine, notwithstanding the fact that there may be complex issues as to whether the state is in breach of the Treaty provision.

Matters are, however, more complex when the language is cast in terms of the EU having failed. This may mean that the EU, normally the Commission, has failed to implement a clear Treaty obligation, or legislative duty, because the policy approach chosen was inefficacious or ill-judged. The EU institutions should be held to account in this respect in the same way as national administrations. Suffice it to say for the present, that there is no evidence that failures of this kind are greater than when policy is determined and applied at national level.

The language of EU failure or blame is, however, commonly deployed to capture something very different. It signals the conclusion that the EU as a collective entity has not achieved something that it should have achieved. When used in this manner it signifies the EU's inability to agree on a desired conclusion, such as a fair and equitable scheme for the distribution of asylum seekers, or a fiscal union based on sharing of burdens through Eurobonds.

The EU should be held to account in this respect, but we should nonetheless acknowledge that the failure is most commonly because the Member States are unable to agree how to resolve the problem, as exemplified by rejection of national quotas proposed by the Commission for asylum seekers. This in turn will often reflect conflicting Member State views as to the limits of solidarity flowing from EU membership. It is important to interrogate such issues, since they lie at the heart of what the EU means as a political and legal order. It is also important when doing so to ensure that our thinking across different areas coheres. Thus, we should consider the relationship between heterogeneity that underpins differentiated integration, and commonality that supports arguments for sharing of burdens in the eurozone or in relation to asylum.

We should also be cognizant, when deploying the language of collective failure, of what the Treaty requires, allows, and prohibits, thereby echoing the point made in the previous section. If we are to cast the language of collective blame we should think carefully whether the failure resides in not doing something that the Treaty demands; whether it entails not taking action that is allowed by the Treaty provisions, albeit not mandated; or whether it signifies the need to amend the Treaties so that action that does not currently fall within their remit can be addressed at EU level in the future. There may be disagreement as to which of these is applicable in a particular case. This can be accepted. It does not thereby obviate the need for such inquiry. These distinctions matter since the normative impulse differs markedly depending on which of the three characterisations best describes a particular situation. Having come to a conclusion on this issue, we should not moreover forget that the Member States are the principal architects of the primary Treaty provision thus interrogated.