How the Court’s path dependence affects its role as a relational actor

Abstract

This Article draws on research into precedent and the European Court of Justice to argue that it is distinctive in almost never retreating from a standpoint it has taken, or overturning an earlier judgement, by contrast with other Supreme Courts where this is a more common occurrence. The Article then considers the implications of this finding for research into the relationship between the Court and other actors, such as Member States, litigants and institutions. It suggests that in considering how the Court may be influenced and constrained this research takes insufficient account of its apparently limited capacity to change doctrinal direction. Evidence of doctrinal path dependence needs to be a more central part of discussions of the Court as a relational actor.

1. Introduction

Discussion of the relationship between the Court of Justice and other actors, or external influences such as public opinion, presumes that to some extent the parties on each end of that relationship are able to respond to each other. For example, the Court may, some think, back down in the face of political pressure, or the Commission or European Parliament may be inspired to actions by judicial decisions.

These interrelations are possible because of the capacity for change, or to put it another way, the existence of choice. It is the fact that the Court and its interlocuters can choose different courses of action that makes it possible for them to influence each other. One cannot have a relationship with a fixed object.

However, a little-discussed but important fact about the Court is that it almost never changes its mind. In the history of its jurisprudence the instances of a decision which contradicts the doctrine of an earlier decision are very rare – far rarer than in the case of its closest peer, the American Supreme Court, and probably also than in the case of apex courts of the Member States of the European Union (EU). Examples where the change in doctrine is towards a more restrictive reading of the law, rather than a more expansive one, are exceptionally scarce. The Court does not retreat.

That fact limits its relational capacity. It could change its mind, certainly, but it appears that it does not. This note suggests that this needs to be integrated more into the relational literature. The theory underlying claims that the Court does or will respond to other actors needs to take account of its apparently limited capacity for change.

The note proceeds in four steps. The first is a brief overview of the range of research on the Court’s relationship to other actors and external influences. The second offers a typology of change, an indication of the different ways in which judicial doctrine can develop, in order to make clear the one specific way in which the Court’s doctrine does not. The third suggests that this is more than just a desire for consistency, which most courts have, but something distinctive to the Court. Finally, a conclusion considers briefly the implications of this lack of change for research on the Court as a relational actor.

2. Relational literature

The idea that courts, and in particular the Court of Justice, are influenced by factors other than pure doctrine is now relatively orthodox, and the categorisation and quantification of that influence has resulted in a significant body of legal and social science research. There is no obvious a priori limit to the scope of this research, because the essential insight is that judges are human, and so their decision-making processes are embedded in the full range of interactions that humans have with the world around them. From quasi-biological issues such as diet and tiredness, to more psychological or sociological ones such as ethnicity, personality and age, or not-yet-explored factors such as the weather and the architecture of courtrooms, the possibilities for exploring influences on judicial behaviour are endless.

However, where the Court of Justice is concerned the literature has focussed on political and institutional factors. A significant body of largely quantitative political science research has explored whether and to what extent the Court responds to Member State interventions, concluding broadly that the Court will not contradict a significant enough coalition of Member States.¹ Extrapolating this political sensitivity, other research has argued that the Court also responds to public opinion, limiting its judgements if there is a risk that they would conflict too harshly with the public mood.² The image is of a court which is concerned not to sail too directly into political headwinds, for the sake of legitimacy and institutional self-protection.

By contrast, many legal scholars – some of whom have challenged the legal assumptions in the scholarship above – and political scientists working with other theoretical frames, have tended to see the Court as more autonomous, an engine driving the bus of integration – or at least of EU law – forward largely whether the Member States want it or not.³ These scholars tend to locate the primary tool of Member State resistance to integration in forms of non-compliance, rather than in containment of the Court. They have also tended to see the Court as an important, sometimes even dominant, influence on the institutions in the legislative process.⁴

Another set of important interlocutors for the Court is national supreme courts. Here it is largely legal scholars who have formulated the idea of the judicial dialogue, and discussed the way in which the Court and supreme courts respond to each other, if not explicitly then at least visibly, adjusting their doctrines in what appears to be response to judgements by the other.⁵ The judicial dialogue is often portrayed as something with a significant element of confrontation, as both sides insist their principles are non-negotiable.⁶ However, a layer of compromise reveals itself upon further investigation, with creative new paths and doctrines being formulated to accommodate the red lines that the other side claims to have.⁷ The suggestion is that the Court listens to national supreme courts, and if it does not concede to them, it takes them seriously and tries its best to find face-saving resolutions to apparent clashes of law.⁸ The essentially relational concept of pluralism has been a prominent feature of EU constitutional thinking generally.⁹

Yet another branch of literature concerns legal mobilisation, or what is also called strategic litigation.¹⁰ As a hierarchically superior and disruptive element in national legal systems, EU law offers opportunities for national lobby groups to leverage domestic change via lawsuits, something that they enthusiastically, if carefully, do. Implicit in this very idea of mobilisation is that the Court can, with the right case, be steered. Yet equally important is the idea that the Court is predictable, and so in a sense that it has a resistance to influences. However, mobilisation sees the strategic litigant as someone capable of producing a legal outcome, not merely a passive element in the legal process. In that sense, there is the idea of a litigant-Court interaction where the power is not all on one side, and so the mobilisation literature could also be said to be relational. The Court is, as with the branches of research mentioned above, an object upon which, and with which, can be worked.

3. The types of doctrinal change

The literature on the Court as a path-dependent actor is in a sense a mirror of the work above. It emphasises the Court’s consistency, and its status as an autonomous actor that can set a course and stick to it independently.¹¹ There is no necessary contradiction between claims that the Court is autonomous, consistent and path-dependent, and claims that it is a relational actor, because neither need to be absolute. Indeed, the very idea of the Court as a relational actor implies both autonomy and responsiveness, the latter being necessary for the relational quality, and the first being necessary for it to be an actor, rather than merely a tool of others. The question of where the Court steers, and where it is steered, can sustain a nuanced and complex answer. Nevertheless, lawyers in particular have tended to emphasise the dominant role of the Court in its relationships, and as part of this sense of the Court as an independent actor have adopted the idea that the Court is loyal to precedent and not easily deflected from its path.¹² This has given birth to a wide range of studies of what that precedent entails, and how it is used.¹³ Yet it has also been argued that while the Court tells a powerful story about the consistency of its own case law, the particular character of precedent in EU law still allows for change and development.¹⁴ The doctrine of precedent is a legitimating factor – a tool to construct an image of a stable and predictable legal system – as much as a genuinely constraining one.¹⁵

There is a disconnect between this literature and the relational research discussed above.¹⁶ That relational research is premised on the idea that the Court can respond and therefore one would expect that the work on how and to what extent it allows itself to change would be an essential part of the theoretical basis. It tends, however, to be ignored. The aim of this note is to show how these two bodies of work could be, and should be, connected to each other. To do that, below there is a typology of the way in which doctrine can be changed. The aim of that is to highlight the one specific way in which the Court does not in fact change its mind. This will clarify the implications for research on how the Court responds to influences from without.

Four types of doctrinal change are considered: technical adjustments, clarification, progression, and restrictive overruling. Judgements in the first two groups are seen as reflecting a need to ‘tidy up’ the law, dealing with uncertainties, incoherencies and loose ends from previous judgements. They do not reflect a judicial choice to take the law in substantive new directions or a change of vision about what the law should be. By contrast, judgements of the second two types do reflect a choice to take a new policy step and change the substantive scope and reach of EU law. Progressive judgements do that in a positive sense, expanding EU law. Restrictive overruling, by contrast, takes the law backwards, contradicting an earlier judgement in order to restrict the law’s scope. It is the holy grail of the Eurosceptic individual or Member State, seeking to tame the law and the Court. It is also the type which is striking in its absence from the law.

These types are not necessarily mutually exclusive, and nor are their boundaries precisely defined. Quite a few judgements could be arguably put in several of the boxes. However, they provide a terminology to describe doctrinal change.

A. Technical adjustments

It is a feature of courts, and particularly of the Court of Justice, that they will try to found their specific decision in general principles. The risk is that in formulating those principles they lay down a rule which has wider implications, and sometimes those implications are hard to see. A well-intentioned statement may turn out to have problematic, or even ridiculous results. For example, in HAG I, on the notoriously tricky interaction between trademarks and the internal market, the Court addressed the question whether a trademark registered in one Member States could be used to keep goods out which bore a similar, or even the same trademark registered in another Member State.¹⁷ Both respect for intellectual property and free movement are enshrined in the Treaty, and it had to decide what to do when the two collided. The Court developed the doctrine of ‘common origin’ which amounted to answering the question with ‘sometimes’ rather than with a ‘yes’ or ‘no’. A few years later the issue came up again, and Advocate General Jacobs wrote a careful opinion explaining that while there was something to be said for both yes and no answers, the compromise approach taken in HAG I led to indefensible distinctions and was logically contradictory and hard to reconcile with any reasonable policy goals. The Court implicitly agreed, because it modified its position in HAG II, dropping the ‘common origin’ doctrine.¹⁸ Reading the judgements and opinions it is clear that this was not an ideological about-turn on the status of trademarks or free movement. Rather it was recognition that the judgement in HAG I, an honest attempt to solve a difficult problem, just did not work in practice.

Technical adjustments may involve direct contradiction of earlier judgements.¹⁹ They may also have policy importance in the field in question. However, they are not driven by power struggles or ideological conflicts. They are just part of the search for coherence and practical solutions. There are not many examples of the Court needing to make the kind of adjustments that it did in HAG II, but for the purposes here, considering the issue of whether the Court can be pushed in certain political directions, they are not very important.

B. Clarifications

Clarifications are similar to technical adjustments, and could even be said to encompass the former category, but the idea of a clarification is somewhat broader. Very often the Court makes statements that are somewhat ambiguous and invite as many questions as are answered. The first case establishing a new doctrine often leads to a flurry of follow-ons in which national courts essentially ask ‘but what about this?’ or ‘did you mean that?’ This is almost the default mode of any legal system based on case law, and one can trace this kind of development from almost any of the major cases in EU law. A recent and notorious example is Ruiz Zambrano, where the Court ruled that Member States were prohibited from refusing residence or work permits to the third-country parents of minor children who were Union citizens, because then those children would have to leave the Union too and so be deprived of the ‘genuine enjoyment of the substance of the rights’ of a Union Citizen.²⁰

What exactly did this mean? What does ‘genuine enjoyment’ encompass? Is there an underlying principle that a Union Citizen cannot be compelled to leave the Union? But when does such compulsion exist? A series of cases followed Ruiz Zambrano, gradually refining its broad statements down to more-or-less workable rules on when a third-country family member can be refused residence rights and when they cannot – albeit that there is still some clarificatory way to go.²¹ There was no contradiction or overruling in any of these cases: on current doctrine, they would all still be decided the same way. However, in response to new situations and questions the Court was able to expand and refine and clarify what it meant.

The line between mere clarification and substantive doctrinal change is not always clear.²² It is the essence of a clarification that an earlier judgement left space for more than one interpretation, and the later judgement then makes clear which of these is right. For adherents of the one not chosen, the latter judgement may seem like a substantive change. Nevertheless, in many judgements the Court quite clearly considers itself to be just explaining an already established doctrinal standpoint, and this, and the degree of change, does allow for a distinction between such cases and ones with major developments of the law. An important question to ask is whether, in the light of the later judgement, the earlier case would still be decided in the same way. That indicates mere clarification.

C. Progression

The progressive development of the law is the beating heart of EU law.²³ All legal systems change and develop, but few will have the deep and evident commitment to a vision of the law as an ongoing and forward-moving project that is explicit in the case law of the Court. This results in a body of law that is fast-changing in many ways, and certainly far less static than national or even international law. There is a constant stream of judgements providing new doctrines, principles, and developments in the law.

In one sense, this can involve overruling. Where the Court may once have denied that a situation fell within the scope of the Treaty, or that a right was available or a remedy was possible, at a later date it may then say that in fact it is within its jurisdiction, or it can be done.²⁴ Szabados, in one of the most recent and systematic overviews of Court deviations from precedent, discusses a number of examples, such as Brown, and Adouille and Corneille, which expanded the protection given to individuals, where earlier judgements had been more restrictive.²⁵ There are also examples in procedural law of the Court moving from limited procedural protections to more expansive ones.²⁶ There are many more examples where a new doctrine is not in contradiction to any previous judgement, but merely to what was previously assumed: before Ruiz Zambrano it was generally considered that EU law would not apply to Citizens in their home states; before Rottmann it was generally considered that Member States were free to determine their own nationality law.²⁷ The judgements overturned prevailing orthodoxy, but did not require any overruling of previous case law.

What all these situations have in common is that they involve the expansion of the scope of EU law, and typically of the scope of its protections and rights. They add another brick to the edifice. In some cases that appears to be because of recognition that an earlier judgement was not working well, but in other cases it is a recognition that circumstances have changed – that time has passed, or new legislation has been adopted, or the EU itself has expanded.²⁸ The Court occasionally refers explicitly to the current stage of integration or of the development of the law, either to explain why it is not extending a right – the EU is not ready for that yet – or to justify developing a new one – what was impossible once, is possible now.²⁹ In both types of situation however it is recognising that EU law is a dynamic and expanding system. For this reason, progressive, expansive, changes of doctrine belong in a category of their own.³⁰ It is not disputed that they are widespread and important.

D. Restrictive overruling

By contrast, restrictive change is very hard to find.³¹ There are almost no cases where the Court overturns an earlier decision and moves to an interpretation of EU law which decreases its scope, or reduces the rights or protections of individuals.³² This was my instinct as a lawyer at the beginning of this research, and I tested that instinct as follows: there is considerable research on precedent in EU law, and its value and methodology.³³ In that literature, cases which explicitly or even implicitly overrule an earlier judgement would be of central importance. A survey of that literature ought therefore to reveal quite easily the extent of such overruling. If the extent was considerable, and the number of examples large, it would then take further investigation to sort through them and see which can be seen as purely technical adjustments, and which are progressive, and which are restrictive. However, it turned out that there is a widespread consensus that restrictive overruling is a rare phenomenon, and the same few examples are repeatedly discussed, albeit that there are several borderline cases where it is arguable to which category they most properly belong.

The first example of the Court overruling itself is HAG II.³⁴ Above this was put in the category of technical adjustment, because HAG II can be better understood as an attempt to solve practical problems resulting from HAG I than as a new vision on the law, but it could be seen as genuine restrictive change.

However, undoubtedly the most obvious and explicit example of restrictive overruling is Keck, which took a step back in the previously wild-west expansion of Article 34 TFEU.³⁵ Its practical consequences are limited, and it is an open question to what extent it even represented a substantive change to the kinds of national measures that the Treaty would disallow, or whether it was primarily a change in the conceptual framework for defining such measures.³⁶ Nevertheless, the Court admitted to having gone too far, and retreated, and that made Keck one of the most discussed cases in EU law.

Also in goods, the Court appeared to retreat from an expansive reading to a narrower one on horizontal effect, first in Dansk Supermarked and then in Vlaamse Reisburos, although it must be commented here that the reasoning in both cases is sufficiently obscure, and the topic sufficiently complex, that it is not fully clear what either case actually rules and therefore whether there was any real change.³⁷ It is probably the case that on similar facts Dansk Supermarked would still be followed, so that this is not really an example of overruling, but the contrary is arguable.

As well as this, there is a well-documented saga of cases concerning Citizenship, at the centre of which are Dano and Alimanovic, which some have considered to represent a restrictive turn by the Court after the radical and foundational Grzelczyk.³⁸ However, while Dano and Alimanovic may have a different mood from Grzelzcyk, they do not overrule it.³⁹ Rather, these cases show the importance of distinguishing between a failure to launch, and a retreat. It is not unusual that the Court introduces a new doctrine, and scholars embrace it enthusiastically and extrapolate all kinds of further possibilities. Then in a follow up judgement, the Court simply stands still, refusing to go further. That is often presented as the Court moving from expansion to restriction, but it is legally more precise to say that it is moving from expansion to stability. Having stated a principle, lawyers try to push it further, and inevitably the point comes where the Court says ‘no, this is far enough’. That is not retreating, but simply a product of the fact that no right is unbounded – the limit will come.

The decisions in Grzelczyk, and somewhat similarly in Ruiz Zambrano, another expansive Citizenship judgement, were greeted with excitement, and then those in Dano and Dereci, two follow-ons, with disappointment, but the latter two cases did not – it is suggested – change an iota of the doctrine expounded in the first two: they simply refused to take it even further.⁴⁰ The first two remain good law. There are discussions to be had about whether the Court ought to have been more adventurous or not, and about whether its reluctance to go further was influenced by Member States or other political factors, but what it did was come to a halt, not go back. These are not overrulings.

At most, these limiting cases could be said to represent a change in trajectory. After one or more expansive judgements the Court will draw its line, and word its judgement in a way discouraging further expansionary argument. As befits this, the rhetoric and tone of these judgements is restrictive and defensive after the grand and open phrases of the earlier progressive decisions. From the perspective which sees EU law as an endless forward motion, these judgements comprise a change in direction. Nevertheless, it remains the case that the earlier judgements are not overruled in this type of case. From the perspective of EU law as a body of law, or doctrine, they are stasis. For the purposes of the argument here, that is the only point that needs to be made.

The Court’s human rights doctrine is another field where overruling might be sought. There is no doubt that in the Solange-inspired series of cases which develop EU human rights law, the Court does change its mind at least partly in response to pushing from the German Federal Constitutional Court (FCC), even if an explicit concession to this effect is not found.⁴¹ Moreover, given that the FCC understood human rights as a necessary constraint on the EU, the Court’s concession to this could be seen as a step restricting the scope of EU law.⁴²

On the other hand, that ‘concession’ has undoubtedly developed over time into a significant expansion. In retrospect it is apparent that the FCC pushed the Court into taking EU constitutionalisation a step further, and that admitting human rights to EU law was an important element in its expansion and development.⁴³ It is a complicated story, but ultimately belongs in the category of progressive change. If the Court did not come with a human rights doctrine until it was pushed, it was probably because it did not dare to be so creative without an excuse – which the FCC thoughtfully provided. At any rate, it may be noted that having recognised the existence of human rights constraints within EU law, the Court has since moved not an inch on the related issues of primacy, direct effect, and judicial competence to review EU actions.⁴⁴ Once the constitutional structure was in place, it was fixed.

Taricco II is sometimes put forward as a counter-example, being the first and only case in which the Court concedes that a fundamental right protected in national constitutional law might lead to EU law being disapplied, whereas in Melloni and Taricco I it had rejected this possibility.⁴⁵ On the other hand, the move in Taricco II is made not by contradicting the earlier judgements, but by recognising that the right in Taricco II was also part of EU law. This was a case where the Court was being pushed by the Italian Constitutional Court, and it could be argued to show the Court allowing national constraints to limit EU law. It does however involve no formal doctrinal overruling – more a spirit of compromise, and an interpretation of existing doctrine that keeps the peace.

At any rate, this is about it. As can be seen, there is space for debate about what is precisely a restrictive overruling or not, and there are also a few more minor counter-examples to be found in literature, which would be considered here to be either technical adjustments or clarifications. But however one measures the situation, there are no more than a handful of restrictive overrulings, and no more than two – Keck and HAG II – that have a good claim to be both (i) restrictive, (ii) an explicit overruling of previous judgements, and (iii) a substantive policy change.

4. Is the Court more path-dependent than other Courts?

Of course, all courts strive for consistency.⁴⁶ To say that the Court has overruled itself a handful of times begs the question of what one might expect: what would be a ‘’normal’’ number of breaks with the past? Looking at other courts could help with this.⁴⁷

The American Supreme Court, in some ways the closest equivalent to the Court of Justice, is certainly much more inclined to change its mind and contradict earlier judgements. A couple of explicit and clear overrulings per year is the average there over recent decades, and indeed over the longer term. Even if one only considers overrulings where the overruled case was from 1957 or later, to make the comparison more balanced, the number of explicit overrulings is between 60 and 100, depending which source one looks at.⁴⁸ That can be compared with fewer than 10 at the Court of Justice.

The comparison is muddied by the fact that the EU is a progressive project, and this note distinguishes between progressive judgements, and overrulings. By contrast, the US Supreme Court has not committed to a federal project, and stands between the federal and state governments in questions of power and competence, at times moving towards one and at times towards the other. A judgement extending the scope of federal powers, where an earlier judgement did not do that, would be an overruling, whereas the equivalent in the EU, where the Court found that it was time to extend EU law a step further, would not be classified in this paper, or generally in the scholarship, as a break. One might perhaps argue that overrulings in the US should be compared with all innovative EU judgements, whether restrictive or progressive.

On the other hand, the progressive nature of EU law is a genuine feature, and a genuine difference between the systems. In the overwhelming majority of innovative, progressive, Court judgements there is no contradiction with an earlier case, but simply a new situation which has arisen, and a correspondingly new piece of doctrine which is developed. The free movement of capital began with cases on movement of money, and then came cases on taxation, and then on government restrictions on investment. Each wave added something to the law and made the law on capital more far reaching and impactful. However, there was no contradiction: nothing in the early cases on suitcases of cash implied that free movement of capital did not encompass golden shares or restrictions on investment – it had just not arisen yet. In this sense, the distinction between innovations and breaks is real and important,⁴⁹ and it is only true conflicts and overrulings which are comparable to overrulings in the SCOTUS.

Given that, it becomes, perhaps unsurprisingly, evident that the US Court is far more inclined to change direction – albeit that it is still a small minority of its judgements when it does. However, turnarounds are an important part of its jurisprudence, and the question of whether it will make them and whether it ought to is a staple of US legal commentary. By contrast, no-one expects the Court to retreat on its basic doctrines, albeit that there is no shortage of those who think that it should.

A comparison with national supreme courts in Europe would entail similar considerations, and be perhaps more enlightening, given that legal culture would be more comparable. Here, unfortunately, numbers do not seem to be available – perhaps part of that European legal culture resists reduction of the legal process to quantified outcomes. However, the idea that a continental constitutional court might take a new approach, substantively or in its form of reasoning, from that which was in earlier cases, is not considered strange by national lawyers. Legal certainty is a value, but so is the independence of each court considering a case before it, including independence from judges which considered cases in the past.⁵⁰ It is accepted that courts change, their composition changes, society changes, and when a case comes to be judged it will be considered afresh, not simply squeezed into the jacket of the past.

This all supports the view that the Court is distinctive in its absence of self-contradiction. It has a direction of travel from which it does not deviate, and when a step it taken it does not retreat.⁵¹ This can be described as path dependence, and both the existence of this high degree of path dependence and its relative neglect in the relational scholarship are striking and important facts.

5. Why is the Court so path-dependent?

The reasons why the Court so rarely overturns its own decisions are largely beyond the scope of this note, although a few remarks may be appropriate.⁵² These reasons are discussed to some extent in the path-dependence literature, and a prominent and plausible explanation is to do with the need to build a new legal system and establish status and legitimacy by showing consistency. As well as this, the Court’s institutional context might contribute to a high awareness of its role not just as an adjudicator, but as one of the parties constructing European integration.⁵³

Alongside these contextual factors – relational factors, one might say – the specific character of EU law may play its role, and not just the remarkable dedication expressed in the Treaties to a mono-directional process of deep integration. Rather, EU law is often noted to have many common law-like characteristics, and to give case law an important, but distinctive role.⁵⁴ The Court cites itself repeatedly, and literally, and often goes to great lengths to emphasise that it is merely repeating what it has already said before.⁵⁵ But despite this almost common-law like dedication to jurisprudential consistency, there is little of the discursive or analytic character of common law judgements,⁵⁶ and no dissenting opinions, both of which provide opportunities to express nuance, disagreement, and other points of view.⁵⁷ These facilitate change, by giving other directions voice in the law, so that future overrulings have a focus to build on: they are not breaking with the past, but just re-reading it and re-ordering its components. By contrast, the Court typically reduces its past judgements to quotable mottos. These may allow for organic, apparently continuous, development of the law, but in their reduction of cases to what can seem like banal truisms they provide little basis for the principled adoption of a new direction.⁵⁸

Another reason, which arises specifically from cases which were investigated for this article, is that it appears that the Court is rarely asked explicitly to change its mind. An examination of a selection of cases which were subsequent to an earlier expansive decision suggests that Member States were often clearly unhappy with the principle established in the earlier case, but that they did not actually ask the Court to reverse it. Rather, they argued that the specific facts of the latter case justified a different approach and should be distinguished.⁵⁹ That, if successful – which it invariably is not⁶⁰ – has the potential to limit the impact of the earlier decision, but does not get rid of it. It is one step short of a full confrontation, a somewhat weasily avoidance of the substantive claim that the earlier decision was wrong.

Whether the descriptions of the Member State arguments in the judgements and Advocate General’s opinions are full and fair, and whether these cases are representative of the case law as a whole, is a question for further research. However, these examples do invite the reflection that perhaps, subject to further research, Member States, and referring judges, are aware that asking the Court to overrule earlier judgements has a low-to-zero chance of success, so that dissatisfaction is better expressed in more oblique ways. That in turn may put less pressure on the Court to overrule. A highly path-dependent Court may create a self-reinforcing dynamic in which it becomes efficient for litigants to work within that path dependence rather than confront it, going for the achievable rather than the optimal, and thereby making the path dependence stronger. That self-limitation by litigants could in turn be important for understanding, and modelling, of how the Court responds.

It must be emphasised that this is a hypothesis for further research, but the point here is that while the outcome put forward in this chapter – that the Court does not change its mind – may be robust and clear, the causative factors behind this may well be complex and diverse.

6. Conclusions

The Court’s failure to change its mind, or to overturn doctrine once it has established it, is of crucial importance for understanding it as a relational actor. For those who study the influence of the Member States or public opinion, their theoretical frame will be implausible and their empirical research inadequately theorised if they do not take account of this limitation. For most often, when pressure comes from these outside sources it is pressure to move away from the progressive, pro-integration path. Certainly, the Court can pause, and it can choose not to move forward but to stay where it is,⁶¹ but it needs to be understood that this is likely to be a temporary victory: the direction of travel is consistent. Moreover, it should be part of such research that the chances of getting the Court to actively retreat, and adopt a new restrictive interpretation when a previous one was more progressive, are almost nil. It almost never happens. The power of outside influences is overstated if this is not taken into account.

For those who study judicial dialogue, similar comments could be made. The Court does not bend on doctrine. If it is going to adapt to national courts that will not come via principled rethinking, for this it does not do. It will come instead via creativity, the use of existing doctrinal bricks to build new structures, as in Taricco II.⁶²

For those who study mobilisation, the point to be extracted is surely that the Court is highly path-dependent, so that the capacity of activists to use it depends very much on what they want to achieve. If they are trying to push EU law forward, they may well be able to ride the judicial horse. If – which is less likely, but might be the case – they want strategic litigation to limit EU law, they can better find other pathways.

There has been a failure of communication between branches of scholarship. Relational studies have tended to proceed as if the game is open, and the nature of the relationship between the Court and other actors is merely a question of power and politics. Scholars of precedent have noted the Court’s consistency, but their work has been used more to understand the mechanisms and principles of case law than to put it in wider context. Allowing these fields to talk to each other suggests new avenues, both theoretical and empirical, not yet explored.

Finally, the Court’s limited capacity for response has another kind of implication for its relations. It means that the degree to which EU law develops is to a greater degree in the hand of national judges and implementing institutions than it would be if the Court was freer.⁶³ As the political mood and the socio-economic context change, pressure for adjustments of the EU legal framework may become intense, and if the Court cannot satisfy these then national courts and implementing institutions will be more inclined to meet the need themselves, using the discretionary interpretative space that they have to increase, reduce, or just change the way that EU law impacts.⁶⁴ In their “dialogue’’ with the Court of Justice, faced with dull and repetitive answers, they may fill in the conversational gaps themselves. In what it is tempting to imagine as a zero sum situation, national users of EU law will occupy a more central and creative place in the development of EU law because the apparent master of that law, the Court of Justice, has chosen to allow itself to be hamstrung by its past.