The new judicial federalism: the evolving relationship between EU and Member State courts

Abstract

A wind of change is blowing through the European judicial landscape. During the past decade, the European Union (EU) judiciary has undergone its biggest institutional overhaul in generations, the rise of authoritarian populism in Central and Eastern Europe has prompted a Rule of Law crisis, several supreme and constitutional courts challenged the supremacy of EU law, while the Court of Justice re-articulated the scope of the duty to refer under Article 267 TFEU and, for the first time, found that domestic last-instance courts breaching it triggered state liability. This Article argues that these and similar developments, once looked at together, suggest that something fundamental has shifted in the EU’s judicial architecture. A new form of judicial federalism has emerged, which departs from the traditional way in which relations between EU and Member State courts used to be structured. Although this new federalism is multifaceted and is marked by both centripetal and centrifugal forces, its distinguishing feature is a stronger centralisation, which manifests itself in a considerably expanded federal judiciary, a greater emphasis on hierarchy, a more careful use of European judicial resources, as well as tighter supervision of national procedures and court structures.

On 5 May 2020, the Bundesverfassungsgericht sent shockwaves through Europe when it announced that the Public Sector Purchase Programme – at that point, the most expansive instrument of monetary policy in European Union (EU) history – violated the Basic Law and, as a result, lacked legal validity in Germany. According to the Karlsruhe judges, the European Central Bank had overstepped its competences when adopting the Programme, and the Court of Justice of the European Union (CJEU) had done a poor, nay, ‘incomprehensibly’ poor job at policing this transgression.¹ Although extraordinary in terms of its political and symbolic impact, the ruling was not the first of its kind. In the decade preceding it, several national supreme and constitutional courts, including in the Czech Republic, Denmark, and Italy, had challenged the authority of the CJEU, or threatened to do so. Nor was it the last. In the wake of PSPP, Polish and Romanian courts have defied the judges on the plateau de Kirchberg in similar terms, if for different reasons.

The growing pushback against the Court of Justice and the supremacy of EU law is but one sign that the European judicial landscape is changing. The past years have witnessed a number of major developments in relation to the operation of EU and Member State courts, as well as the relationship between them. The EU judiciary has undergone its biggest institutional overhaul in generations, resulting in the elimination of the Civil Service Tribunal and a doubling of the General Court, which may soon receive jurisdiction to decide in certain types of preliminary reference proceedings. The rise of authoritarian populism in parts of Central and Eastern Europe has prompted a Rule of Law crisis, with governments implementing reforms that curtail the power and independence of judges. Several constitutional courts chose to submit their first ever preliminary references to Luxembourg, overcoming their historic reluctance to participate in the European judicial dialogue. At the same time, the Court of Justice re-articulated the scope of the duty to refer under Article 267 TFEU and, also for the first time, found that national last-instance courts breaching it had triggered Member State liability.

These and similar developments are typically discussed separately; the present Article will connect them or, perhaps more accurately, argue that they are connected by analysing them through the lens of judicial federalism. Federalism has been a popular concept in studies on the political structure of the EU.² Yet, as elsewhere in the world, scholarship on European federalism has primarily focused on its legislative and executive aspects. Judicial federalism, ie the question of how judicial power is divided in multi-level government structures, remains less well explored. With few notable exceptions,³ the concept has so far remained off the EU constitutional lawyers’ radar.⁴ This is not to say that there has been little academic engagement with judicial matters in the EU. The role of and interaction between the European and national judiciaries has yielded a rich body of literature, including on each of the topics just mentioned. Yet, the underlying phenomena tend to be analysed in an isolated manner. Scholarship on institutional reforms is largely detached from that on the acte clair doctrine which, in turn, has only limited overlaps with that on the supremacy of EU law. The added value of using judicial federalism as a theoretical framework lies precisely here. It allows – even forces – us to look at the judicial system as a whole. The focus on the allocation of judicial power links jurisdictional, procedural, and substantive questions, thus giving us a more in-depth understanding of the functioning of a legal order and the role which its different components are playing.

The objective of this Article is to examine the evolution which the EU’s judicial system has undergone. It will be argued that the developments which have been unfolding over the past few decades, but especially over the past ten years, suggest, once looked at together, that something fundamental has shifted in the European judicial architecture. A new form of judicial federalism has emerged, which departs from the traditional way in which relations between EU and Member State courts used to be structured. This new federalism is multifaceted and marked by both centripetal and centrifugal forces, yet its distinguishing feature is a stronger centralisation. While the various instances of pushback against the supremacy of EU law have dominated public perception, raising fears of a disintegration of the European Gerichtsverbund, there have in parallel been significant and, ultimately, more influential efforts to strengthen EU judicial authority. They manifest themselves in the expansion of the EU judiciary, a greater emphasis on hierarchy, a more careful use of European judicial resources, as well as a tighter supervision of national procedures and court structures.

The changes have an impact on the enforcement of European law. The growth of federal judicial capacities enables the EU to solve a larger number of disputes concerning federal (EU) law centrally. The intervention in national procedural autonomy and the organisation of Member State judiciaries increases control over how EU rules are implemented at the national level. But there are also more subtle, no less important repercussions for the relationship between European and national courts. The insertion of vertical elements into the preliminary reference procedure forces domestic judges, notably those at the top level, to participate more actively in the dialogue with the CJEU, a move away from the idea of cooperation towards that of hierarchy. At the same time, national courts are increasingly interpreting EU law autonomously and are nudged by the Luxembourg judges to decide cases of limited reach and importance on their own. The result is an altered division of judicial functions which has knock-on effects for the uniformity of EU law.

The Article proceeds as follows. It will begin by introducing the concept of judicial federalism, outline the various shapes it can take, and explain what is commonly thought to define the EU’s own brand of federalism. Subsequently, some of the central developments that have taken place in the court structure of the EU and its Member States will be examined. These will be grouped into two categories: centralising and decentralising tendencies. Finally, some reflections will be offered on the reasons for these changes and the implications they have for the European integration process. It will be argued that the shifts at the judicial level mirror the broader evolution of the EU towards a fuller, more mature polity, suggesting that the fate of judicial and political federalism may be intertwined to a greater extent than is commonly assumed.

1. Judicial federalism(s)

At its core, federalism denotes the idea of self-rule and shared rule: political authority is partly exercised by local governments, partly by the central government.⁵ It is traditionally contrasted with, on the one side, the concept of a unitary state in which sovereignty comes from one source only and, on the other, that of confederalism in which states cooperate on certain matters through an international agreement but retain ultimate sovereignty themselves. Federalist polities emerge when independent states enter into a union with one another or when unitary states devolve powers to sub-national authorities. It is widely accepted that the EU, despite lacking the quality of a state, is built on the principles of federalism.⁶ Its powers of government are divided between the Union and the Member States – or, in the language of federalism, between the federal level and the constituent units, or the centre and periphery.

Federal entities often (not always⁷) have a judicial structure that mirrors the division of functions at the political level. It can entail a division of the judiciary into federal and state courts, as well as procedures for policing the federal compact. The rationale behind such mechanisms is two-fold. For one, they are meant to ensure that both federal and state governments stay within their allocated scope of competences;⁸ it is federal overreach that has proven more problematic in practice.⁹ In addition, they are meant to ensure that federal law is respected and applied with a sufficient degree of uniformity across the polity. This is a non-trivial challenge given that in many federal systems federal law is, if to different extents, enforced by state courts. This creates the potential for a local misapplication of, or even resistance against, federal legislation and adjudication. Further, since there is, at least usually, a substantive number of state courts, there is a danger that the way in which federal law is interpreted or applied across the states will diverge, thus weakening its effectiveness.

Underneath these high-level similarities, there are important differences in the way in which judiciaries are organised across federal polities. Three main types of judicial federalism have emerged over time: unitary, dual, and integrated.¹⁰ In unitary systems, there is no division of the judiciary into federal and state courts. Instead, a single judicial structure exists which is responsible for settling disputes under both federal and state law and is typically administered by the federal government. Belgium falls into this category, as does South Africa. Dual systems have separate court structures at the federal and state level. Depending on their degree of complexity, this can include a full network of lower, appellate, and supreme courts for each level. In their purest version, dual systems divide tasks strictly – federal courts enforce federal laws, state courts implement state laws – but, in practice, there tend to be overlaps and control mechanisms between the two judiciaries. The United States, Argentina, and Switzerland exemplify this approach. Integrated systems are a half-way house between the unitary and dual models. They have federal and state courts but, instead of being kept in separate spheres, both are embedded within the same hierarchical structure. Decisions of state courts can be appealed before the federal judiciary which, as a result, tend to be smaller than in dual systems (as a substructure of lower courts is rendered superfluous). Germany and Canada have adopted this integrated model.

It bears noting that the foregoing are just ideal types. Even within these categories, we find substantial differences.¹¹ In addition to the elementary question of having a single, dual, or integrated court system, there is a number of structural choices every legal order must make which end up affecting the particular shape of its judicial federalism. The most visible example are rules on jurisdiction. Having multiple courts or even levels of courts in a judicial system raises the issue of competence. There are numerous ways in which judicial competences can be allocated, ranging from the legal rules invoked, to the substance of the subject-matter, to the origin of the litigant, each resulting in a different separation of tasks between federal and state courts. Relatedly, the availability of appellate review will determine whether, when, and by whom a given court’s judgements can be re-examined. But there are also less immediately obvious factors that can impact on the division of judicial authority. Rules on how judges are appointed, the size of the different parts of the judiciary, as well as issues such as which level of government has the competence to regulate procedures are all relevant in this context. Each of these choices will contribute to the eventual level of centralisation or decentralisation of the federal arrangement, producing systems in which the federal and state judiciary play, respectively, a stronger or weaker role.

A comparison between the US and the EU is instructive in this regard. Although both have federally organised judiciaries, the US model is marked by a higher degree of centralisation.¹² It follows the dual approach, with two parallel court structures at state and federal level, each organised and financed by the respective government. State courts primarily handle litigation relating to state law. In addition, there is a network of federal courts which are responsible for disputes concerning federal law. And this network is large: there are 94 first-instance courts (District Courts), 13 appellate courts (Courts of Appeals), and a Supreme Court. The result is that claims concerning federal law tend to be brought before and decided by federal courts.¹³ When state courts adjudicate on federal matters, they are subject to federal oversight. (The same holds true, albeit to a more limited extent, when they adjudicate on state matters.¹⁴) Despite the existence of safeguards protecting its jurisdiction and interpretive authority,¹⁵ the state judiciary is ultimately in a position of subordination vis-à-vis the federal judiciary. It is bound by the supremacy clause, under which the US Constitution and all federal legislation as well as treaties made pursuant to it must form the ‘supreme Law of the Land’, and must abide by the interpretation of that body of law propounded by the US Supreme Court.¹⁶ Judgements of state supreme courts touching on federal law can be appealed before the US Supreme Court which, however, has certiorari jurisdiction, meaning that it enjoys practically unfettered discretion as to which cases it takes on and which ones it refuses.¹⁷

The EU has embraced the idea of judicial federalism, too, but has opted for a system with more decentralised traits. Judicial functions here are also divided, between EU and Member State courts. Each level of government is, in principle, solely responsible for the establishing and organisation of its courts, as well as for laying down the procedural framework within which they operate. Yet, the system is marked by a notable asymmetry. Whereas we have fully developed court structures at the domestic level, the EU judiciary consists only of the Court of Justice, a system resembling integrated federal models. Consequently, the large bulk of adjudication is done by the (more numerous) national courts. EU courts are competent to interpret and assess the validity of EU law, without exerting, at least formally, any control over the validity of Member State law. National courts have a dual function: as parts of the domestic judicial system, they adjudicate issues of national law; as juges communs of the EU, they apply European law in disputes brought before them.¹⁸ There is, unlike in the US, no explicit hierarchy between the two judiciaries. The Court of Justice does not have the direct power of appellate review over rulings of national judges. Instead, both levels are connected through the preliminary reference mechanism enshrined in Article 267 TFEU. It enables national courts – in the case of courts of last resort, it obliges them – to submit questions concerning EU law to the Court of Justice, where an answer is necessary to solve a domestic dispute. It creates a system based on ‘cooperation’ and ‘judicial dialogue’.¹⁹

Comparing federal systems in this way can be useful as it demonstrates the large variety of possibilities when it comes to organising the judiciary in composite political structures.²⁰ It exposes the respective advantages and disadvantages, and it can help us to understand what impact certain attributes of a system have on judicial authority, the enforcement of federal laws, and the protection of rights. However, focusing on inter-systemic variation carries the risk of overlooking another important aspect of federalism: intra-systemic evolution. Juxtaposing different legal orders to identify similarities and differences makes it necessary to rely on ‘snapshots’ of how a given system was constituted or what it has come to look like. As a result, the picture that is painted constitutes a mere reflection of a specific time and context.

In federalism studies, there has recently been a movement, spearheaded by scholars such as Patricia Popelier, towards emphasising the dynamic nature of federal structures.²¹ Traditionally, federalism is thought of as a fixed, unchangeable compact that allocates powers between the different levels of government in a specific way and protects each side against interferences from the other.²² Yet, reality shows that federalism is not a static phenomenon; it evolves over time.²³ The relationship between centre and periphery within a certain polity, as encapsulated in a certain institutional settlement, can and typically does change. Constitutions are amended, competences are granted and withdrawn, new institutions are created, and old ones begin to exercise their roles differently. All this can lead to shifts in the ‘federal balance’.²⁴ These are not just limited to political aspects of federalism. They may also affect its judicial dimension.

Against this background, a dynamic reading of the EU’s judicial federalism will be proposed in the following. It will be argued that, within the basic structure of the European judicial system that was outlined above, a series of critical changes have taken place during the past decades, but in particular during the past ten years. The developments pull in two different directions: towards a centralisation and decentralisation of judicial authority. Centralisation and decentralisation were introduced earlier as categories allowing us to understand the differences between federally organised judicial systems, as markers of how much power the federal and state level respectively hold. This power can manifest itself in a variety of ways, including the question as to whether a state or federal judiciary exists at all, the jurisdiction it has been granted, the circumstances under which its decisions can be appealed, and so on. The same can be true within a single judicial system. Here, too, the degree to which judicial power is centralised, ie lies within the federal level, or is decentralised, ie is in the hands of the states, can evolve. In some periods, the courts of one level of governments may be the dominant force in the legal order; during others, the picture may – partly or fully, gradually or abruptly – reverse.²⁵ To what extent and how the respective power of EU and Member State courts has evolved will be the topic of the next two sections.

2. Centralising tendencies

The EU judicial system has changed over time. Partly these changes have occurred as a result of explicit constitutional or institutional reforms, partly through lower-level policy updates, and partly through a judicial reinterpretation of existing provisions in the European Treaties and Member State constitutions. One group of changes embodies a move towards centralising judicial authority in the EU: the growth of the European judiciary; the increasing emphasis on hierarchy in the relationship between EU and Member State courts; an emerging concern for prioritisation, and greater control of national judicial processes.

A. Growth of EU judiciary

The first, and perhaps most visible, sign of how the EU judicial architecture has evolved concerns the size of the federal judiciary. When the European Communities were established, they were given their own permanent court to help settle disputes arising under the rules of the Treaties, which was remarkable at the time. However, in the beginning, the European judiciary was very small. It consisted of one body only, the Court of Justice, which was composed of seven judges and two Advocates General. They were the sole federal judicial element in what would soon be declared to constitute a ‘new legal order’.²⁶

Over the years, the EU judiciary has substantially grown. With every new Member State joining the European Communities, later Union, the Court of Justice received an additional judge to deal with the increased workload and guarantee national representation in Europe’s fabrique du droit, up to a total of 28 before the United Kingdom’s departure from the Union. The number of Advocates General was periodically increased, too; it presently stands at 11, of which 6 are permanent and 5 allocated on a rotating basis. Yet, and more importantly, the European judiciary has not just grown by what can be considered a proportionate amount based on the given number of Member States – it has expanded far beyond that. The steep increase in litigation made it necessary to expand EU judicial capacities. In this way, new institutions emerged to help with the task of adjudication.

The first sign of this expansion was the establishing of the General Court (GC) or, as it was originally called, the Court of First Instance. In the 1980s, the number of cases brought before the Court of Justice began to reach unprecedented levels, resulting in delays that many felt were intolerable.²⁷ In addition, there was a sense that the EU did not do enough to ensure the judicial protection of individuals living under its laws.²⁸ Consequently, the decision was taken to create a second court that would take some work off the Court of Justice’s hands. The GC’s jurisdiction was limited at first. Substantively, it primarily dealt with competition and staff cases. Procedurally, it only handled actions for annulment brought by private parties. With time, the GC’s competences have been continuously broadened and now include cases on trademarks, state aid, and common foreign and security policy. Its procedural jurisdiction covers actions for failure to act, actions for damages, and arbitration proceedings. Decisions on questions referred for a preliminary ruling are, as will be discussed below, likely to be added in the future.

In 2005, a third judicial body followed. The Treaty of Nice opened up the possibility of creating ‘specialised courts’ within the EU’s judiciary,²⁹ an option which was made use of for creating the Civil Service Tribunal. The Tribunal was established to deal with disputes concerning EU staff matters, an area that had begun to burden the General Court. And it discharged this task, according to a universal consensus, efficiently.³⁰ Despite its success, the Tribunal was dissolved after only a decade as part of a wider overhaul of the European judiciary.³¹ The key component of the reforms, which were meant to provide a remedy to the increasingly unmanageable docket of the GC, was an enlargement of the GC at the expense of abolishing the Civil Service Tribunal (CST).³² Decried as a horse trade,³³ the changes resulted in a further enlargement of the EU adjudicative process. The 7 judges that were lost as a consequence of the dismantling of the Tribunal were replaced with 28 additional judges at the GC (one less post-Brexit).

Even in its current, much expanded shape, the EU judiciary lags behind other federal judiciaries in terms of size and complexity. Unlike in dually organised countries like the United States or Argentina, there is no full federal court structure with a network of trial and appeal courts, which would complement the judicial systems of the Member States. The differences with integrated systems also remain pronounced, with the federal judiciary being significantly smaller than, for instance, in Germany or Canada.³⁴ Nonetheless, the expansion of the EU judiciary over the years has generated a significantly higher capacity to resolve disputes concerning EU law through its own courts. The European Court now has just under 100 members. It receives a combined total of almost 2,000 disputes per year.³⁵ The EU judiciary temporarily had, while the CST was in operation, a veritable three-tiered system of judicial review, in which rulings of the Tribunal could be appealed before the GC, whose decisions could then be reviewed by the Court of Justice. Due to the Tribunal’s dissolution, and until another specialised court is established,³⁶ one tier has been removed from this structure. However, a three-tiered system de facto continues to exist in areas in which individuals can lodge complaints before an independent board of appeal first (eg against decisions of the European Union Intellectual Property Office or the European Chemicals Agency), can appeal the board’s decision before the GC, and have the latter’s ruling reviewed before the Court of Justice.

This division of functions also means that the EU judiciary is no longer as top-heavy as it used to be. In most countries there is an inverse correlation between a court’s rank in the judicial hierarchy and the number of cases it decides. For a long time, the opposite was true in the EU. The ‘highest’ court, the Court of Justice, was also the busiest, its case load exceeding that of all ‘lower’ courts (GC and CST) combined. The emergence of additional judicial bodies and a gradual broadening of their responsibilities – as well as the phenomena of hierarchisation and prioritisation – have led to a reversal of this relationship. During the past decade, the adjudicative output of the GC has overtaken that of the Court of Justice, a trend that is likely to deepen in the future.³⁷ As a result, the GC has, to an important degree, become a real lower federal court.

B. Hierarchisation

A second aspect of the EU’s changing judicial federalism is the increasing hierarchisation. Traditionally, European law has shied away from the concept of hierarchy. As explained, there is, formally, no relationship of superordination and subordination between the EU and Member State judiciaries. The Court of Justice has been eager to emphasise that the two are bound by the principle of cooperation, with each actor contributing in different ways to the common objective of administering justice in the EU. In some ways, the European integration project has even contributed to actively subverting judicial hierarchies. The principles of direct effect and supremacy, paired with the duty of judicial review of Member State acts,³⁸ have relocated power in national legal systems, strengthening the authority of lower courts at the expense of higher courts. Academic observers have long called for (re-)introducing elements of hierarchy into the European judicial architecture – for reasons of principle as well as in order to manage the CJEU’s growing case load.³⁹ But it is only since the turn of the century and, more seriously, during the past few years that we see that idea materialising. This notably shows in the way in which the role of national supreme and constitutional – or apex – courts has evolved within the European judicial architecture.

The first symptom of this evolution concerns the use of preliminary references. As part of their relationship of cooperation with the Court of Justice, national courts are, for one, meant to apply EU law in domestic proceedings when it is relevant to a dispute, and do so correctly. For another, they are meant to participate in the preliminary reference mechanism laid down in Article 267 TFEU. The responsibilities connected with this duty depend on the place of a court within the domestic chain of command. Lower-ranking courts may refer questions concerning the validity and interpretation of European law to the CJEU (unless they plan on invalidating an EU act⁴⁰), last-instance courts must.

For a long time, there was a curious asymmetry between the letter of Article 267 TFEU and the way in which the system operated in practice. While lower courts were eager to participate in the European judicial dialogue, most national apex courts refused to submit preliminary references to the CJEU.⁴¹ This has begun to change. Persistent objectors like the Spanish Tribunal Constitucional, the French Conseil Constitutionnel, and the German Bundesverfassungsgericht chose to make their first references in the 2010s.⁴² A study conducted by Pavone and Keleman shows that this is symptomatic of a broader trend: references from national supreme and constitutional courts have increased over the past two decades and overtaken those submitted by lower-tier courts.⁴³ This is not to say that there is no room for further engagement.⁴⁴ Yet, it suggests that apex courts have finally understood that the preliminary reference system is not just a threat to their authority but can also potentially strengthen their influence by allowing them to communicate with, and potentially impact on, a wider European judicial audience. The Court of Justice appears to incentivise this volte-face. It is privileging references from supreme and constitutional courts by declaring them admissible more frequently than those submitted by lower courts,⁴⁵ thus putting back in place elements of the national judicial hierarchy that EU law had so famously eroded.

This, if you will, is the carrot, but there is also a stick. The Court of Justice is not just facilitating that national apex courts participate in the preliminary reference mechanism, but it increasingly demands that they do. For decades, it had been an open secret that many supreme and constitutional courts refused to seek the Court of Justice’s help with questions concerning EU law, instead interpreting the relevant rules on their own. The EU tolerated this behaviour for a surprisingly long time, but it does no more, at least not to the same extent. Köbler laid the foundations for change.⁴⁶ Drawing on its Francovich jurisprudence, the Court of Justice held that actions for state liability could be brought where a decision of a court of last instance infringed EU law, notably by manifestly breaching the Court’s case law. In the EU, there is – unusually for a federal judicial system – no way of appealing a judgement of a Member State court of last instance concerning European law. Therefore, violations of EU rules by apex courts can, in principle, not be remedied. Köbler introduced, as Komárek noted, an ‘indirect possibility to appeal’ by allowing litigants to initiate a review of a national judgement before the domestic judiciary.⁴⁷ This is not the same as a reversal, but it does create a legal and economic sanctioning mechanism which effectively forces national apex courts to comply with CJEU rulings, a decisive step away from the idea of cooperation towards that of hierarchy.

Despite its theoretical importance, Köbler remained a dead letter for many years. Scholars put this down to the high threshold for liability that the Court set in the judgement.⁴⁸ (Francovich claims have proven difficult to enforce in general.⁴⁹) This changed with Ferreira da Silva e Brito.⁵⁰ A dispute brought by the collective redundancy of airline employees ended up before the Portuguese Supreme Court, which was asked by the applicants, but denied, to make a preliminary reference on the interpretation of the Transfers of Undertakings Directive, alleging that the relevant provisions were clear. The applicants challenged this ruling by means of a state liability action before a lower domestic court, which chose to seek advice from the CJEU. The European justices found that the Supreme Court had not only misapplied their case law, it had failed to submit a reference due to erroneously deeming the issue to be clear, thus triggering the liability of the Portuguese Republic.⁵¹ For the first time, a judgement of national apex court was successfully challenged for violating EU law.

The developments surrounding Köbler liability are best read alongside those relating to the acte clair doctrine. In CILFIT, the Court of Justice introduced an exception to the otherwise categorically worded Article 267(3) TFEU.⁵² Domestic courts of last instance were exempted from their duty to refer in situations in which the correct application of EU law was ‘so obvious as to leave no scope for any reasonable doubt’. If the Court of Justice gave national judges an inch, they took a mile. CILFIT was abused by uncooperative domestic courts, which now had the possibility to refuse to submit a preliminary reference under the guise of fully complying with EU law; Turmo’s contribution to this special issue provides an illustrative case study.⁵³ It has been pointed out that the strict conditions imposed in CILFIT have, inadvertently, contributed to that state of affairs. The CJEU asked judges to ensure that the meaning of the relevant EU rule was clear not only to them but also to courts in all Member States, which requires to compare each language version of the norm. Against this backdrop, AG Wahl quipped that ‘coming across a “true” acte clair situation would, at best, seem just as likely as encountering a unicorn’.⁵⁴

The wide-spread practice of avoiding references via the acte clair route, too, were silently accepted by the EU for a long time, but there has been a change of heart recently. The aforementioned Ferreira da Silva e Brito was a first warning shot. It was followed by Commission v France, in which the Commission successfully brought an infringement proceeding concerning the Conseil d’État’s failure to make a reference in a tax case.⁵⁵ There were temporary indications that the Court was, prompted by its Advocates General, considering a more lenient reading of the acte clair doctrine.⁵⁶ Yet, the recent Consorzio Italian Management signals a different path.⁵⁷ While slightly relaxing the CILFIT criteria in so far as the comparison of different language versions and national jurisprudence are concerned,⁵⁸ a change that is likely to be of more theoretical than practical effect,⁵⁹ the Court has also imposed a new obligation on last-instance courts: they must now provide a statement of reasons whenever they choose not to make a reference.⁶⁰ The ruling constitutes an attempt to improve compliance with the duty to refer by increasing the costs of non-compliance. It is also a message to domestic apex courts that, even if they enjoy certain liberties within the preliminary reference mechanism, participating in it – and, thus, respecting the authority of the CJEU as the final arbiter of EU law – is not a voluntary act, but a mandatory feature of the European judicial system.

C. Prioritisation

In parallel, and to some extent as a result of the hierarchisation process, we have also started to see a growing concern for prioritisation at the EU judiciary. By prioritisation, I mean an effort to direct one’s attention towards issues that are the most significant for the development of the legal order and, just as crucially, away from those that are of minor importance only. The need for prioritising arises when demand for adjudication exceeds judicial resources – and judicial resources are famously scarce.⁶¹ Despite their expansion, EU courts have struggled for decades, if to different extents at different points in time, with processing all disputes brought before them. This has led to the development of strategies concerning how many cases they take on, which ones, and how these are decided.

One area in which this prioritising approach shows is within the EU judiciary itself. The Court of Justice used to take on every dispute that was thrown its way and decide, especially during the first phase of its existence, as a plenum. There were no pressing reasons to do otherwise. Throughout the 1950s and 60s, as well as during the first half of the 1970s, the case load was manageable, at times even positively low. As legend has it, champagne was served in the corridors of the palais de justice when the first preliminary reference from a Member State court arrived. This changed quickly. Due to the combined effect of the enlargement of the European Communities, the activist jurisprudence of the Court, and the growing familiarity of national judges with European law, case numbers began to grow and create capacity issues. Faced with this challenge, the Court of Justice put a chamber system in place to render its decision-making more efficient. Parts of the adjudication would be done by the full Court, parts by smaller judicial units.

The chamber system has, over time, been expanded and employed with ever growing enthusiasm. In its current format, which was introduced in 2012,⁶² cases are, as a rule, assigned to chambers of three or five judges. Only where the ‘difficulty or importance of the case or particular circumstances’ require it, or a Member State or EU institution which is party to the proceedings requests it so, will a case be assigned to the Grand Chamber, which consists of 15 judges.⁶³ (In exceptional situations, the Court of Justice can also sit as a full court.⁶⁴) The result of this structure is that the vast majority of disputes are nowadays settled by chambers, whereas only cases which are significant for the development of EU law are dealt with by the Grand Chamber.⁶⁵ This creates a stark division of functions within the Court of Justice. The chambers practically serve as a lower judicial structure that does the mundane adjudicative work. The Grand Chamber, in turn, has become a real apex organ inside the Court which sets the general directions of its jurisprudence and decides the high-profile constitutional cases. This role is, both in qualitative and quantitative terms, not too different from that played by institutions like the US Supreme Court.⁶⁶

A prioritisation process can also be observed in the way in which the EU’s appeal system has developed. Every multi-tiered judicial structure faces the question of how to regulate access from the lower to the higher echelons. When the GC was created, the right to appeal to the Court of Justice was formulated very widely. Parties and privileged interveners were, in essence, automatically entitled to appeal on any point of law. This became untenable over time as case numbers before the GC – and because of that, numbers of appeals before the Court of Justice – increased. To counter this trend, the Rules of Procedures introduced the possibility to dismiss appeals as manifestly inadmissible or unfounded.⁶⁷ The Court has been making use of that option more and more aggressively over time in order to limit its case load. In relation to cases that were already considered by an independent board of appeal and the General Court, a 2019 amendment to the Rules of Procedure has further restricted the scope for appeals. Such disputes had been a thorn in the Court of Justice’s eye for a while because they came in great numbers and tended to be without merit. Now, an appeal will only be allowed to proceed ‘where it raises an issue that is significant with respect to the unity, consistency or development of Union law’.⁶⁸ This is a consequential reform as it practically gives the Court of Justice full autonomy in relation to which questions it wants to pronounce itself on – a certiorari power of sorts.⁶⁹ Again, the effect will be that only matters of doctrinal, political, or constitutional importance will reach the top of the EU’s judicial system, whereas the rest will be left for the lower judiciary.⁷⁰

The latest and perhaps most significant development in this area is the announcement that the General Court may soon obtain jurisdiction to rule on preliminary references. Currently, preliminary references fall within the exclusive purview of the Court of Justice. Although the Member States had created the possibility for the GC to be given the power to ‘hear and determine actions or proceedings referred for a preliminary ruling.. . in specific areas’ already in the Treaty of Nice,⁷¹ this route lay dormant until now. Too strong was the CJEU’s reluctance⁷² to share its jurisdiction in relation to what has been described as ‘the jewel in the Crown of the existing regime’.⁷³ The Court’s request from November 2022 to amend its Statute signals a sea change.⁷⁴ It proposes a transfer of jurisdiction to the GC concerning preliminary references in six areas – cases on VAT, excise duties, the Customs Code, passenger rights, and greenhouse gas emission allowance trading – insofar as these ‘do not raise questions regarding the interpretation or validity of Union law of a horizontal nature’.⁷⁵ It appears that, on the one hand, the continuous rise in litigation before the CJEU and, on the other, the successful enlargement of the GC made the Court re-consider its stance. The reform would mean that disputes on low-profile but high-frequency topics, on which a well-established jurisprudence exists, could be delegated to the GC. As a result, the CJEU’s bandwidth to resolve more complex and relevant legal issues would increase.

A growing concern for prioritising is also beginning to show in the way in which the Court of Justice structures its relationship with Member State judiciaries. A seemingly technical matter symbolises this: inadmissibility rulings. During the first two decades of its activity, the Court of Justice was very receptive to preliminary reference requests from national judges, no matter how poorly phrased, irrelevant, or insignificant they were on occasion. This was the phase of docket building and the establishing of the EU legal order, and the Court did not want to scare national judges away as it understood that they were an important ally. Consequently, almost every preliminary reference was accepted as admissible and decided on the merits. As the case numbers started to grow, this approach became hard to sustain. So, the Court began – cautiously – to set boundaries as to what types of questions national judges could ask. In a series of rulings in the late 1970s and early 1980s, it established that references involving artificial disputes,⁷⁶ hypothetical problems,⁷⁷ and questions unrelated to European law⁷⁸ were inadmissible. A little later, it did the same for references that fail to provide sufficient information on the relevant factual and legislative context.⁷⁹ The use of inadmissibility rulings remained sporadic for a while. Since the 2000s, they have considerably increased in number⁸⁰ and doctrinal relevance.⁸¹ Although this development has affected references from all types of national courts, it has, as noted, been particularly prominent regarding those submitted by lower-instance judges. The Court of Justice’s stricter approach appears to be motivated by the desire to create capacity, and to use this capacity for things that matter (more) from a perspective of EU law.

A similar shift is occurring in other areas. A myriad of procedural tools have emerged which, in different ways, all pursue the same objective: to allow the Court to dispose of cases faster and in a less resource-intensive manner. One of these is the possibility of deciding without a prior opinion from the Advocate General where the proceedings raise no new point of law.⁸² Another innovation is the option of deciding by means of reasoned order, ie without having to write a full judgement. In a recent study, Šadl et al found that such orders are increasingly popular with the European judges, who use them strategically to deal with sudden influxes of cases in a policy area or to ‘dodge certain debates that arise due to local and legally narrow problems’.⁸³ This procedural strategy is flanked by substantive efforts such as the use of deference doctrines, through which the CJEU delegates decisions it does not feel the need to make itself to national courts.⁸⁴ Again, we see a concern for not squandering valuable judicial resources on issues which are of minor importance for the EU legal order, but using them where they are truly needed.

Despite their significance, it is important not to overstate the magnitude of these developments. The Court of Justice is more protective of its docket and embraces doctrinal and procedural tools that enhance adjudicative efficiency. However, it also continues to decide many run-of-the-mill disputes itself – too many, as most commentators, including from within the Court, would say.⁸⁵ Overall, there is still a tangible difference between the CJEU and other federal apex courts when it comes to the level of involvement in the interpretation and application of federal law. In part, this is due to the legal and institutional framework in which the Court operates. The preliminary reference procedure especially, which creates a direct path to Luxembourg for national courts with relatively few constraints, puts pressure on the system. In parts, it is also of the CJEU’s own doing. While in some areas the Court has started to exercise interpretive self-restraint,⁸⁶ in others it engages in exceedingly detailed factual adjudication, rendering its own work taxing and indirectly creating further demand for references.⁸⁷

Nevertheless, it is clear that the growing prioritisation changes, step by step, both the Court of Justice’s internal functioning as well as its external relationship with national courts. What might be less clear is whether this is, indeed, a sign of centralisation of judicial power in the EU: does it, after all, not contribute to reducing the number of cases decided by the CJEU? I shall discuss this issue below in greater detail but, suffice it to say at this point, there is, indeed, an important element of decentralisation in the just-described developments. The greater care with which the Court of Justice curates its docket has the effect of pushing certain decisions down to the GC and national courts. This might, at first glance, look like a weakening of the Court’s authority. Yet, by deliberately passing on the possibility to rule in certain circumstances, the CJEU is, perhaps paradoxically, strengthening its position as the central arbiter of the most fundamental legal issues in the EU. Less can, at times, more: by rising above some of the more mundane problems, the Court of Justice acquires more gravitas within the system.

D. Control of national judicial processes

A final development pointing in the direction of greater centralisation is the EU’s interference in national judicial processes. The European judicial architecture rests on the idea that the Member States, in principle, have autonomy over the organisation of their judiciaries. That extends to decisions relating to the jurisdiction of courts, the appointment of judges, rules of procedure, and any other aspect of the judicial system. Yet, over time, the Court of Justice and, to some extent, the EU political process have been asserting more control over domestic choices here. This becomes evident when looking at two areas which may, on the face of it, appear unconnected: the case law on national procedures and the measures taken to address the Rule of Law crisis.

The question as to whether and how EU law impacts on national remedies and procedures has fascinated scholars to such an extent that it has become, by now, its own field of inquiry. I shall limit myself to a brief overview of the milestones in this area. The starting point is the principle of national procedural autonomy. In two landmark rulings in the 1970s, the Court of Justice declared that it was for each domestic legal system to designate which court has jurisdiction and to determine the procedural conditions governing legal (as well as administrative) actions intended to ensure the protection of EU rights.⁸⁸ However, it qualified this autonomy in two ways. The first is the principle of equivalence, which requires that Member States do not discriminate between actions concerning domestic rights and EU rights. The second is the principle of effectiveness, which stipulates that national procedures do not make the enforcement of EU rights impossible or exceedingly difficult.

It is the principle of effectiveness that has proven particularly consequential in practice. The Court of Justice has used it to impose a variety of obligations on Member States, ranging from the need to introduce new remedies, to allow state liability claims for violations of EU law, to effect changes in relation to more prosaic matters like the burden of proof, time limits, and legal aid.⁸⁹ The Court’s activity in this area has gone through different periods.⁹⁰ The 1970s and early-‘80s were marked by a deferential approach, which left considerable leeway to the Member States. They were followed by a decade of interventionism that culminated in the activist case law of the early-1990s. After that, the Court appeared to revert back to a slightly more cautious stance,⁹¹ but this did not last. Since the 2000s, there has been a return to interventionism,⁹² with the principle of effectiveness – as well as, increasingly, Article 47 of the Charter⁹³ and the principle of sincere cooperation⁹⁴ – being used more enthusiastically to limit national procedural autonomy. A recent study conducted by Wallerman shows that, between 2008 and 2019 alone, the Court of Justice has struck down over 40 national procedural rules (although part of the fault might lie with integration-friendly national judges).⁹⁵ Add to this that there has been a fair deal of sectoral EU harmonisation in relation to procedures and remedies⁹⁶ and you understand why some commentators believe that there is, in reality, no such thing as a principle of national procedural autonomy anymore.⁹⁷

In addition, the CJEU has started to control more tightly how national judiciaries are organised through its Rule of Law jurisprudence. Here, unlike in the domain of procedure, the focus is not on the precise way in which disputes involving EU rights are resolved, but on ‘macro issues’: the composition of domestic courts, their relationship with other branches of government, and their status within the legal order more broadly. The Portuguese Judges case was an early harbinger of this stream of case law.⁹⁸ As part of the austerity measures relating to the Euro crisis, Portugal reduced the salary of employees in the public sector. The reforms, inter alia, affected the judges at the Court of Auditors, who brought an action against the salary cut, claiming that it infringed the principle of judicial independence by robbing judges off their financial stability. It was unclear whether there was merit in this argument, but it was even less clear whether the issue fell into the scope of EU law at all. The Court of Justice ruled that it did. Relying on the second subparagraph of Article 19(1) TEU, which was introduced in the Lisbon Treaty and obliges Member States to ‘provide remedies sufficient to ensure effective legal protection in the fields covered by Union law’, and Article 47 of the Charter, which enshrines the right to effective judicial protection, the Court held that protecting the judicial independence of national courts was essential. The concept of independence presupposed that a court ‘exercise[d] its judicial functions wholly autonomously’ and was ‘protected against external interventions or pressure’ that could influence the decisions of its members.⁹⁹

The salary cut in Portuguese Judges did not fall foul of these requirements, but the combination of Article 19 TEU and Article 47 of the Charter has become a powerful tool which allows the Court to influence how national judiciaries are structured and operate. Its full potential has been revealed during the Rule of Law crisis. Over the past decade, several Member States in Central and Eastern Europe have experienced what is commonly referred to as a turn to illiberalism,¹⁰⁰ democratic backsliding,¹⁰¹ and, in some cases, a ‘constitutional breakdown’.¹⁰² Governments in these countries – with Poland and Hungary leading the way – have adopted a range of reforms that are aimed at curtailing the authority of courts. After initial hesitation, the EU decided to intervene to protect the Rule of Law, with the Court of Justice showing particular determination to push back against the destruction of Member State judiciaries. The constitutional discourse accompanying this intervention usually centres on the need to protect fundamental European values.¹⁰³ Yet, the Court’s activity in this area has also had, perhaps to some extent unintended, consequences for judicial federalism. By entering the Rule of Law space, the Luxembourg judges have expanded their control over the organisation of Member State courts.

The case law in this area has, by now, reached vast dimensions. It is explored in greater detail in the contribution of Bornemann, so I shall limit myself to sketching the essentials. Much of the CJEU’s attention has been focused on the situation in Poland. The Court declared several prominent government initiatives to be in breach of EU law, voicing concerns that actors within the Polish judiciary were insufficiently protected from external political influences. This includes the reform of the ordinary judiciary, which entailed a lowering of the retirement age of judges but gave the Minister of Justice the power to authorise extensions;¹⁰⁴ an analogous reform of the Supreme Court, which retroactively lowered the retirement age of its justices while granting the President of the Republic unfettered discretion to prolong appointments;¹⁰⁵ the establishing of a special Disciplinary Chamber which is competent to reprimand judges for the content of their decisions;¹⁰⁶ and changes to the composition of the National Council of the Judiciary which aimed at bringing it under closer legislative supervision.¹⁰⁷

The EU’s involvement in policing the Rule of Law continues to expand. More and more cases, from within and beyond Poland, are reaching the Court of Justice. A successful challenge was brought against rules subjecting lower judges in Romania to disciplinary liability for failing to comply with decisions of the constitutional court.¹⁰⁸ A similar Hungarian mechanism that penalised national courts for submitting preliminary references was found to violate Article 267 TFEU.¹⁰⁹ The appointment process for members of the judiciary in Malta, in which the Prime Minister enjoys significant discretionary powers, was reviewed, but ultimately given the green light.¹¹⁰ More recently, the EU political process has come to the aid of the Court. Instruments such as the Rule of Law Conditionality Regulation and Next Generation EU exert financial pressure on Member States to respect basic standards of judicial independence and fair trial. Although questions have been raised as to whether they will be able to effect meaningful change,¹¹¹ they deepen the EU’s enmeshment in the organisation of Member State judiciaries. European institutions acquire a supervisory role relating to how national courts are composed, judges are appointed, rulings are reviewed, and so on.

3. Decentralising tendencies

The foregoing section explored the centralising trends in the EU judicial system. However, not all changes that have taken place over the past years and decades can readily be understood as falling into this category. There are equally developments that pull in the opposite direction, indicating a move towards decentralisation through a strengthening of national judicial power. Two phenomena illustrate this: the pushback against the supremacy of EU law and the autonomous interpretation of EU law. Both are driven by national supreme and constitutional courts.

A. Challenges to supremacy of EU law

Probably the most controversially discussed aspect of the relationship between EU and Member State courts concerns the supremacy or primacy of European law. Over the past decade, several national supreme and constitutional courts have revolted against EU rules and Court of Justice judgements interpreting them, arguing that these cannot take legal effect in national law. Such instances of pushback, although limited in number, are a symbolic and increasingly also practical threat to the EU’s judicial system.

The authority of EU law is a perennial topic that has occupied lawyers for almost half a century. The story has been told many a time. The Court of Justice established the principle of supremacy in the 1960s, holding that European law derives its authority from the Treaties and takes precedence over national law in cases of conflict.¹¹² The principle is, as subsequent rulings clarified, absolute: any EU rule trumps any colliding national rule, including those of constitutional rank.¹¹³ National courts accepted this finding in general terms, but qualified it in two important ways. First, they were eager to underline that European law only applied in and superseded domestic law because national legal orders allowed it to do so. In other words, its authority was rooted in national, not European law. Second, and partly because of that, there were limits to the principle of supremacy. Here, different national courts took different approaches, yet converged on a set of common themes. Supremacy would be denied in cases in which the EU acted ultra vires, did not ensure a sufficient base level of fundamental rights protection, or interfered with elements of national or constitutional identity.¹¹⁴

For decades, the conflict between the Court of Justice and national courts remained theoretical. Both sides reiterated and further substantiated their positions, but ultimately always ended up on the same page: EU law took precedence over conflicting provisions of national law. This began to change in the 2010s. In Landtovà, the Czech Constitutional Court decided to disapply a Court of Justice ruling on the pension rights of workers from former Czechoslovakia.¹¹⁵ The Czech judges found that the ruling was ultra vires; the Court had wrongly interpreted the relevant EU social security regulation. In Ajos, a seemingly trivial dispute concerning old-age pensions and severance allowances, the Danish Supreme Court disregarded a prior preliminary ruling of the CJEU.¹¹⁶ The judgement articulated a harsh critique of EU adjudicative practice, suggesting that the Court of Justice, by creating unwritten general principles, not only violated the Danish constitution but the whole idea of separation of powers. More cases followed. The Hungarian Constitutional Court put a halt to the application of the EU refugee relocation scheme in Hungary by referring to the country’s constitutional identity.¹¹⁷ The Italian Constitutional Court threatened to go down the same route in Taricco,¹¹⁸ a case concerning limitation periods for the prosecution of VAT offences, but ultimately averted an open conflict by submitting a second preliminary reference and, thus, giving the Court of Justice the opportunity to adjust its case law.¹¹⁹

Although these acts of resistance created a stir across Europe, they were still widely thought of as outliers.¹²⁰ They had an air of idiosyncrasy and mostly concerned smaller Member States at the political periphery of the Union. This changed in May 2020, when the German Federal Constitutional Court rendered its PSPP ruling.¹²¹ The German judges declared the homonymous ECB programme, which had been adopted as part of the EU’s response to the Euro crisis, and a ruling of the Court of Justice that had approved it, ultra vires. They criticised that the CJEU had not done a sufficiently thorough job at reviewing the programme and its proportionality, resulting in an unconstitutional inflation of the ECB’s competences. What is more, the Court’s interpretation of the EU Treaties was ‘arbitrary’ and ‘not comprehensible from a methodological perspective’ – consequently, the CJEU had ‘manifestly exceed[ed] the judicial mandate conferred upon [it] in Art 19(1) TEU’.¹²² The decision was nothing short of a bombshell. The most influential domestic court in Europe, located in the Union’s largest Member State, had blocked a measure of monetary policy with a previously unseen financial volume. After almost 50 years of barking, Karlsruhe had finally bitten.

What was immediately raised as a danger in the wake of the ruling became a reality soon afterwards: courts in authoritarian and euro-sceptic Member States (ab)used PSPP as a blueprint.¹²³ Following the path trodden by the Bundesverfassungsgericht, the Polish Constitutional Tribunal held, inter alia, that a reading of Article 19(1) TEU which would give domestic courts the power to question the legality of judicial appointment procedures was inconsistent with the Polish Constitution.¹²⁴ Despite the evident parallels between the two cases, there were also relevant differences. While the German ruling challenged the validity of a secondary act and the CJEU judgement upholding it, the Polish decision objected to a certain interpretation of primary law while, formally, leaving the CJEU’s jurisprudence intact (at least for now¹²⁵). In the meantime, yet another judicial conflict has emerged. The Romanian Constitutional Court has defied the authority of the CJEU in a prolonged and continuing exchange on the legality of domestic reforms concerning the disciplinary liability of judges.¹²⁶ Directly after the Court of Justice found the legislation to violate EU law and instructed lower courts to disapply it,¹²⁷ the Romanian constitutional justices declared that their decisions remained generally binding according to the Romanian Constitution; any countervailing EU ruling could only produce effects after the Constitution was revised.¹²⁸

It is unlikely that we have seen the last of the struggle for supremacy. More cases, involving different legal issues but also, potentially, courts from different Member States, are bound to transpire.¹²⁹ At this point, it is too early to tell how the confrontations between the European and national judges will end, and which constitutional lessons will be learned from them. The EU, for its part, is gradually changing its approach on this issue. Initially, it had chosen to ignore the first instances of recalcitrant rulings, which meant that they went legally unchallenged. PSPP led to a re-thinking of that strategy. A few days after the German Constitutional Court’s decision had been announced, the Court of Justice published a press release commenting on the judgement – a highly unusual course of action – in which it underlined the binding nature of its rulings.¹³⁰ The Commission initiated infringement proceedings against Germany for failing to respect the supremacy of EU law and the authority of the Court of Justice. Yet, in December 2021, it announced that these were closed as the German government ‘formally declared that it affirms and recognises the principles of autonomy, primacy, effectiveness and uniform application of Union law’ and committed ‘to use all the means at its disposal to avoid, in the future, a repetition of an “ultra vires” finding’.¹³¹ Germany has gotten off with a slap on the wrist. It need not be that other countries will receive the same gentle treatment,¹³² if the phenomenon continues to spread.

While there is a risk of exaggerating the legal and practical impact of these events, it would be a mistake to underappreciate their significance for the EU’s judicial system. There were voices in the literature that downplayed the impact of rulings like PSPP, arguing that instances of national judicial defiance remain the exception.¹³³ In the vast majority of cases, the many thousands of domestic courts would still conscientiously apply EU rules and follow decisions of the Court of Justice. This may (or may not) be true.¹³⁴ Nonetheless, the challenges to the supremacy of European law and the authority of the Court are relevant, especially, but not only, if viewed through the lens of judicial federalism. At a symbolic level, they pose a threat to the delicate system of cooperation and dialogue that is at the heart of the European judicial architecture. Having dispensed with establishing a formal hierarchy and direct system of appeals, the EU, more than other federal polities, is dependent on the faithful implementation of its laws by Member State courts. In this light, having over a quarter of domestic apex courts refusing, or considering to refuse, to apply EU acts and rulings constitutes a serious problem. But even at a practical level, the matter is increasingly causing difficulties. Whereas cases like Slovak Pensions, Ajos, and PSPP did not have major ripple effects in the domestic judiciaries, the situation is different when it comes to the Polish and Romanian rulings. In both countries, we are seeing lower courts follow the lead of their constitutional courts. In this way, the resistance against the Court of Justice is already having real implications for the street-level application of EU law.

B. Autonomous interpretation of EU law

A second development has almost got lost in the flurry surrounding the battle for supremacy, but may have similarly far-reaching consequences: national supreme and constitutional courts have started to interpret European law autonomously. Decisions like Landtová, PSPP, and the Polish Constitutional Tribunal’s Rule of Law judgement all engage, directly or indirectly, in an interpretation of the competence norms laid down in the European Treaties, concluding that their reach is narrower than purported by the CJEU. But there is, in addition, a growing number of instances in which domestic judges assert the authority to interpret substantive provisions of EU law. This phenomenon can notably be observed in relation to fundamental rights.

Fundamental rights have always been a sensitive aspect of European integration. The original Treaties famously lacked any rights protections. This began to trouble national courts in the 1960s, especially those in Germany and Italy, countries which had discovered the significance of having effective human rights safeguards after going through periods of totalitarianism. To fend off challenges to the authority of European law, the Court of Justice created fundamental rights ex nihilo by reading them into the EU’s constitutional materials as unwritten general principles. In turn, national constitutional courts vowed that they would not subject EU acts to domestic rights review. The result was what been aptly described as the ‘separation thesis’, a strict partition between the two legal spheres.¹³⁵ National courts would only assess the compatibility of national acts with national fundamental rights, leaving the constitutional review of EU acts, as well as that of national acts on grounds of EU fundamental rights, entirely to the Court of Justice.

With time, the EU’s fundamental rights acquis has grown. More and more general principles have been recognised and used in an increasingly creative fashion. In 2000, the Union gave itself a proper bill of rights by adopting the Charter of Fundamental Rights. In the Lisbon Treaty, it additionally committed itself to acceding to the European Convention on Human Rights, a project that is still to be finalised.¹³⁶ In this new setting, the Court of Justice decided to update the EU fundamental rights framework. In Akerberg Fransson and Melloni,¹³⁷ it, on the one hand, defined the Charter’s scope of application widely – drawing on its previous jurisprudence on general principles – and declared that national courts could not deviate from EU fundamental rights in harmonised sectors, even if they sought to guarantee a higher level of protection. On the other, as a quid pro quo, it granted Member States a margin of appreciation in not (fully) harmonised areas, allowing domestic standards as long as they did not compromise the ‘primacy, unity and effectiveness of EU law’.¹³⁸

Around the same time, Member State courts began to reconsider their stance on EU fundamental rights. The Austrian Constitutional Court acted as a pathbreaker in this context. In an asylum case, it was asked to decide whether the domestic authorities’ refusal to grant the applicants legal protection violated Article 47 of the EU Charter of Fundamental Rights.¹³⁹ Given that the Court’s competences had always been understood, including by itself, to extend to protecting domestic constitutional rights only, with the Supreme Administrative Court handling EU law-related disputes, it was expected that the claim would be declared inadmissible. Yet, to the surprise of most commentators, the Constitutional Court asserted its jurisdiction, departing from the rule whereby EU law cannot be invoked in constitutional proceedings.¹⁴⁰ Based on a creative reading of the principles of equivalence and cooperation, it held that rights contained in the EU Charter of Fundamental Rights – not, however, those recognised as general principles – constituted a standard for domestic constitutional review insofar as comparable rights existed in the Austrian constitution.

Other constitutional courts followed suit.¹⁴¹ The Belgian Cour constitutionnelle established that Belgian and European constitutional provisions, including those on fundamental rights, constituted an ‘inseparable whole’ and, therefore, had to both be treated as integral elements of judicial review.¹⁴² Italy’s Corte costituzionale found that in cases in which both EU and domestic fundamental rights were at stake, ordinary courts had the duty to first seek the advice of the Constitutional Court before making a preliminary reference to the Court of Justice.¹⁴³ In its rulings on the ‘right to be forgotten’, the Bundesverfassungsgericht departed from its longstanding jurisprudence and accepted EU fundamental rights as a direct standard of review, opening up the possibility for litigants to rely on them in constitutional complaints. ¹⁴⁴ In addition, it held, making use of the discretionary scope opened up by Akerberg Fransson,¹⁴⁵ that German constitutional rights would apply in areas which had not been fully harmonised and, more surprisingly, that there was a presumption that the level of protection guaranteed by the Grundgesetz satisfied the requirements of the Charter.

These developments can be read as an attempt of domestic apex courts to assert themselves. As a consequence of the new line of jurisprudence, they are no longer just guardians of national constitutional rights – they have also become guardians of EU fundamental rights. This constitutes a break with the ‘separation thesis’ which had dominated adjudicative practice for decades, a break that has implications on two fronts. Internally, it leads to a rebalancing of power within Member State judiciaries.¹⁴⁶ The longstanding refusal of national supreme and constitutional courts to engage in EU fundamental rights review did not mean that EU rights were absent from domestic disputes during all this time; it just meant that these courts were largely cut out from implementing them. Individuals invoked EU fundamental rights against acts of both national and European authorities. The proceedings were adjudicated by lower courts who, in the case of interpretive doubts, liaised with the CJEU. By opting to engage with EU rights, national apex courts claim their seat at the table and regain some of the power they had ceded to their colleagues further down in the domestic judicial hierarchy.¹⁴⁷

Externally, the changes affect the relationship between Member State and EU courts. National apex courts no longer leave the protection of EU rights to the Court of Justice but, instead, claim an active role in their enforcement.¹⁴⁸ Most of the aforementioned courts have, explicitly or implicitly, made clear that they will submit preliminary references to Luxembourg when questions about the correct interpretation of EU fundamental rights will arise, especially on matters concerning fully harmonised areas. Yet, this should not, as Thym notes, be misconstrued as an ‘abdication’ – quite to the contrary, we can expect constitutional courts to turn ‘the threat of the “last word”.. . into a forward-looking power of the “first word”’.¹⁴⁹ Instead of being limited to reacting ex post to the CJEU, by taking the extreme step of defying its interpretation of a particular right on domestic constitutional grounds (think Taricco), they can now articulate their own interpretation ex ante. Especially national courts with a high reputational standing and a strong record of rights protection may profit from this as they will, to a greater extent than before, be able to influence the substantive content of EU fundamental rights.

4. Our judicial federalism in flux

The EU’s judicial federalism is changing. Its basic structure, by and large, continues to look like it did during the foundational phase of the European integration project: a system composed of a comparatively small EU and comparatively large Member State judiciary, which share the responsibility for enforcing European law; each level of government enjoying the primary competence for the organisation of its courts; and the lack of a formal hierarchy between the two judiciaries which are, instead, connected via the preliminary reference procedure and the idea of judicial cooperation. However, within these general parameters, significant changes have taken place, in particular over the past decade.

Both centralising and decentralising tendencies can be detected. Although the latter have, through high-profile conflicts like PSPP and the Rule of Law saga, received the lion’s share of public attention, it is the former which, overall, have the upper hand – and quite clearly so. The EU judiciary is getting larger and larger; its relationship with national courts is acquiring more hierarchical streaks; the Court of Justice is becoming more selective about which cases get on its docket; and there is stronger EU intervention in national procedural rules and the organisation of Member State courts. Seen individually, each of these changes may appear of limited consequence. However, taken together, they prompt a tangible relocation of judicial authority from the national to the European level. The result can be described as a greater federalisation of the EU judicial system or, more precisely, a shift in the EU’s federal balance. The centre’s power over judicial matters has grown, the periphery’s has shrunk.

Two reasons may account for this shift, both related to the Union’s federal structure. Many federal polities suffer from a centralisation bias.¹⁵⁰ Over time, the division of political power between the federal and local units gets skewed, with the former gradually absorbing competences of the latter.¹⁵¹ This phenomenon can be facilitated by the federal judiciary, which is usually given the task to enforce the allocation of powers between both levels of government but, as a federal institution, tends to protect the prerogatives of the centre with greater vigour than those of the constituent units.¹⁵² The evolution of the EU judiciary suggests that a similar dynamic can unfold in relation to judicial power. The division of functions within a composite judicial system can evolve and experience a centralising drift. Even more so than in their role as umpires of legislative and executive competences, federal courts may, given their direct stake in the matter, find it hard to resist expanding the scope of federal influence. The consequence is a judicial ‘competence creep’.

Similarly, there may be a certain parallelism in the evolution of judicial and political federalism. The EU started off as an international organisation whose objective was largely limited to creating an internal market between its Member States. With time, its competences have grown. Every Treaty change has brought a broadening of its mission statement, which has come to include areas as diverse as police cooperation, environmental protection, monetary policy, and space exploration. These competence gains have been complemented with symbolic advances such as the establishment of EU citizenship and the articulation of a joint set of foundational values. In this way, the EU went from being a mere economic enterprise to a fully-fledged political project. It appears only logical that its judicial system would somewhat echo this transformation. In a recent article comparing judicial federalism in the EU and US, Wells has argued that, in federal polities, deeper degrees of political integration create a demand for a stronger and more efficient federal judiciary.¹⁵³ The enlargement of EU courts can be seen as a response to this demand. It allows the Union to enforce its laws to a greater extent through its own judiciary. Where it must rely on Member State courts, the intervention in domestic procedures and judicial systems, as well as the insertion of elements of hierarchy, enhance its control over how EU rights are applied.

It bears noting that the new, more centralised type of judicial federalism does not simply mean that EU courts are doing more and Member State courts less; the current federal arrangement is more nuanced. EU courts may, overall, be deciding a larger number of cases than they used to, but national courts are also given wider leeway to adjudicate and enforce EU law autonomously. The Court of Justice can afford – it is, due to the steady increase of workload over the years, even forced – to limit its docket with greater determination. It dismisses a growing share of cases without an examination of the merits, or disposes of them in a resource-saving way by making use of the many procedural tools that have been created for that purpose. In preliminary reference proceedings, it can focus on the dialogue with national apex courts, which hold a particularly important position in the European judicial architecture, and reduce its involvement in matters of minor importance or purely local scope. Similarly, when it comes to appeals from the GC, it can concentrate on addressing the most consequential legal issues.

This has an important impact on the enforcement of EU law. The growth of federal capacity to process disputes involving EU rules contributes to promoting their uniform interpretation and application across the Union. But that uniformity is not spread evenly. The Court of Justice pays great attention to centrally clarifying the most fundamental aspects of European law. Especially the Grand Chamber is gradually embracing what can be called a ‘guiding function’. It has, paraphrasing Chalmers, come to assure ‘the unity and ordered development of [the EU] legal system through setting out a number of steering judgments each year’.¹⁵⁴ At the same time, the European judges show greater restraint when it comes to non-fundamental matters. The Court no longer feels the need to settle each and every legal problem itself. It is prioritising more strongly, leaving certain decisions in the hands of the GC and Member State courts and, thus, creating greater room for diversity in the way in which EU law is applied. In this way, a differentiated approach is emerging, which is defined by high(er) degrees of uniformity on questions of principle and the acceptance of low(er) degrees of uniformity on questions of detail.

Even if materialising only gradually, this change is no minor adjustment. For a long time, the Court of Justice,¹⁵⁵ supported by much of the legal literature,¹⁵⁶ embraced a maximalist conception of uniformity. Any weakening of the uniform interpretation and application of European law was considered a grave danger to the functioning of the EU judicial system. The logical upshot of this view was that only one court in the entire legal order – the CJEU – should lay down the meaning of EU law, with local divergences to be strictly avoided. This conception has been criticised as outdated and harmful.¹⁵⁷ It rests on an unrealistic and, even by national standards, exceedingly rigid understanding of uniformity, which assumes that any deviation from a single, centrally-defined standard will result in the demise of the legal order.¹⁵⁸ More importantly still, it hinders active engagement with, not just against, EU law by Member State courts, by denying them agency in its interpretation and, in this way, turning them into passive by-standers. The developments we are seeing signal a welcome rethink on this front. They show that the EU legal order is maturing and beginning to espouse a division of judicial functions which resembles other federal systems, with a more balanced distribution of duties between the different tiers and levels of courts.

All this has repercussions for the idea of judicial cooperation. For national courts, cooperation is traditionally understood to entail the duty to submit questions concerning EU law to the Court of Justice and loyally apply its jurisprudence. Now, the notion of cooperation is increasingly acquiring a further, more active dimension. National judges must implement and, where necessary, interpret EU law independently to a greater extent than they used to. Where the Court decides against providing a ruling or signals it will not solve a particular legal problem, Member State courts have to fill the resulting gap. This ‘responsibilises’ domestic judges. They are disincentivised to seek the support, or an ‘alibi’,¹⁵⁹ from the CJEU and, instead, nudged towards taking certain decisions autonomously. That way, EU law is developed to a higher degree in national courtrooms.

Despite the predominance of centralising tendencies, there are equally movements in the opposite direction. They, too, are defining features of the EU’s new judicial federalism. One might be tempted to dismiss the push for decentralisation as a failure of domestic judges to read the signs of the times. Yet, that we are seeing national courts reasserting their power at this point might not be a coincidence. To some extent, it reflects the spread of post-functionalist sentiments across Europe, which have generated euro-scepticism among citizens and governments.¹⁶⁰ More importantly, it is the direct result of the centralisation process that has been unfolding within the EU judiciary. Member State courts, especially at the top tiers, sense that domestic legal orders are to an ever-greater degree pervaded by European law and the authority of the CJEU. This seems have triggered a two-sided reaction. On the one hand, it made national apex courts realise the benefits of participating in the European judicial dialogue. On the other hand, it activated their instinct to resist. Therefore, it might be too early to bury the judicial empowerment thesis, as recently proposed by Pavone and Kelemen.¹⁶¹ National supreme and constitutional courts may have become the CJEU’s primary interlocutors, but they have not lost the will to defend and, where possible, expand their powers through alternative means.

Among the decentralising features, the autonomous interpretation of European law by national courts is the less controversial as well as, from a federal perspective, less problematic one. This is not to say that it does not have a significant impact. It effectively takes away the monopoly of the Court of Justice on interpreting EU fundamental rights. Although this might lessen the uniformity of EU law and lead to delays in administering justice, it could ultimately be a small price to pay if the outcome is instilling a greater sense of commitment to the European project among Member State judges.¹⁶² Presuming that domestic courts deliver on the promise of referring preliminary questions on at least some of the issues arising in EU fundamental rights law– and there are, as explained, ways of encouraging them to do that¹⁶³ – we might end up with a modus operandi that reflects how many mature federal judicial systems operate. The Court of Justice will provide guidance on the most fundamental or challenging legal problems and step in where divergences at the national level pose too great a threat for the EU legal order. Much of the day-to-day interpretive work will be done by Member State courts.

The national pushback against the supremacy of European law and the CJEU’s authority is certainly the more delicate aspect of decentralisation. Instances of state courts challenging the authority of the central government or judiciary are not unprecedented; one also finds them in well-established federal systems.¹⁶⁴ Nonetheless, if these become a regular occurrence, as seems to be happening in the EU, they insert instability into the legal order. There are two ways of thinking about the issue. The first is to accept such acts of resistance as a legitimate and necessary feature of the EU’s judicial federalism. Writing before the latest wave of resistance against the authority of EU law, Schütze observed that ‘the normative ambivalence surrounding supremacy and sovereignty’ is ‘part and parcel of Europe’s federal nature’.¹⁶⁵ On this account, the ambiguity about the ultimate locus of sovereignty is not a temporary nuisance, but an essential feature of the federal structure. Therefore, allowing national courts to voice their opposition to EU claims of absolute supremacy is vital as it is the only way of upholding that ambiguity. Conflict is built into the very DNA of (judicial) federalism.¹⁶⁶

A second, more commonly held view is to think of the domestic challenges to EU supremacy as a problem that needs to be overcome. While musings about national limits to the validity of European law may have been tolerable, for some constitutional pluralists even desirable,¹⁶⁷ as long as they stayed within the realm of theory, the situation is different now that words have turned into action. In addition to endangering the successful implementation of EU policies, the disapplication of CJEU rulings damages the unity and effectiveness of EU law, thus putting the European judicial system at risk. It is here that the costs of the type of judicial federalism which the EU has espoused are laid bare most glaringly. Member State courts will – sometimes unknowingly, other times intentionally – breach EU rules. Yet, due to the lack of a system of direct appeal, these breaches cannot be automatically and authoritatively corrected by the Court of Justice. The EU can increase the costs of non-compliance by bringing infringement proceedings and facilitating state liability actions, which can exert both legal and financial pressure on the given Member State. But, ultimately, it is dependent on the good will of national political processes (to amend domestic statutes or constitutions) or adjudicative processes (to revisit their jurisprudence) for remedying violations of European law.

After the PSPP ruling was rendered, analogies were quickly drawn to the political and constitutional conflicts in the antebellum United States.¹⁶⁸ In the first half of the 19^th century, several state supreme courts revolted against the US Supreme Court, refusing to comply with its rulings and challenging its jurisdiction. This was in addition to a number of state legislatures ‘nullifying’ federal legislation. The episode exacerbated the tensions between the North and the South, which eventually led to the Civil War. During the Reconstruction era, the powers of the federal government and judiciary were substantially strengthened. The experience of state defiance, judicial and political, ended up becoming a cautionary tale and an important factor in mobilising efforts for greater centralisation. It is not unthinkable that a similar dynamic, minus the bloodshed, could play out in the EU some day. If the costs resulting from national resistance against the CJEU become too great, and their effects are felt across the Union, this could generate the necessary political momentum for adopting more extensive safeguards to protect EU laws and court rulings. In this way, the push for decentralisation could, inadvertently, trigger further centralisation.

5. Conclusion

This Article looked at the evolution of the EU’s judicial system over the past decades, but particularly over the past ten years. It argues that the structure of the Union’s composite judiciary, which consists of EU and Member State courts, has undergone a series of changes. These changes fall into two categories. The first represents a movement towards centralisation, or a strengthening of federal judicial power. This includes the expansion of the capacity of the EU’s own judiciary, the introduction of elements of hierarchy into its relationship with Member State courts, the growing concern for prioritisation at the Court of Justice, and the greater control exercised over national judicial processes. The second set of changes go in the opposite direction, towards a decentralisation of the judicial system and a reinforcing of the role of domestic courts. The manifold challenges to the supremacy of EU law and the authority of the CJEU are relevant here, as is the autonomous interpretation of EU fundamental rights by national supreme and constitutional courts. Overall, the centralising tendencies decidedly outweigh the decentralising tendencies, leading to a shift in the federal balance from periphery to centre. However, both developments are significant as it is their combination that defines the EU’s new judicial federalism.

The European judicial architecture has not developed in a vacuum. Scholarship on European integration has become accustomed to emphasising the disjointed nature of the EU project. The judicial process has advanced, so the classical story goes, independently from and asynchronously to the political process, which has lagged behind.¹⁶⁹ The findings of the present Article suggest that the parallels between the evolution of the EU’s judicial and political federalism may be stronger than commonly thought. The Union’s legislative and executive competences have continuously grown, its raison d’être has become multi-faceted. It is particularly since Maastricht and, more recently, in the long decade since the Euro-crisis that we have seen significant leaps in integration. The EU has created large-scale financial instruments to counter domestic solvency issues, put in place mechanisms to accommodate extraordinary numbers of refugees, turned into a dedicated regulator of digital services, increased its activity in public health governance, and embraced the idea of debt mutualisation, just to name some of the landmarks. This has generated the need to have a stronger and more effective judiciary, a need to which the centralisation of judicial authority is a response. Going forward, the combination of growing regulatory ambition and growing economic interdependence will make a (more) reliable implementation of European rules indispensable. Consequently, the dynamics between political and judicial federalism are likely to remain important. The more political integration in the EU will deepen, the more its judicial system will have to adapt.