Cilfit Still Fits

Introduction

What is a good case? Although the question may sound bizarre to the ear of most academics, who are primarily interested in what a good (or a bad) judgment is, judges are probably more familiar with it. It is indeed more usual for those on the bench to have a discussion on whether the cases upon which they are asked to adjudicate are suitable for a reconsideration of previous judgments and whether such reconsideration should take the form of a mere clarification, a revisiting, an overhauling or even an overruling. Making such a decision is obviously not easy. It depends on a mix of endogenous and exogenous factors: on one side, the facts of the case, its legal context, the type of legal issues at stake, the capacity of the bench to come up with a new, workable and assumingly better solution; on the other side, the timing, the political and judicial context, the embeddedness or the popular support of a long-standing solution.

Against that background, it is rather clear that the case in Consorzio Italian Management ¹ was not deemed by the Court of Justice of the European Union (referred to throughout as the Court of Justice or simply the Court) to constitute a good case in which to reconsider in depth its early Cilfit judgment ² on the scope of Article 267, third paragraph, TFEU, whereby a national court of last instance shall refer to the Court of Justice any question concerning the interpretation of EU law that is raised in a case pending before it (‘the duty to refer’). In Consorzio Italian Management, the Court of Justice decided instead to stick to Cilfit. In particular, it did not seize the opportunity to overhaul the three famous exceptions to the duty to refer that were laid down in Cilfit: where the question raised is irrelevant; where the EU provision in question has already been interpreted by the Court; or where the correct application of EU law is so obvious as to leave no scope for any reasonable doubt (the famous acte clair doctrine).

However, the case was considered good enough by the Court of Justice to tinker with the exceptions to the duty to refer in two ways: first, by slightly revising the evaluation of the acte clair exception; second, by imposing a new duty on national courts of last instance to give reasons for their decision not to refer in the light of those very exceptions. Those are the visible novelties brought in by the Court. Most interestingly, the judgment in Consorzio Italian Management also contains invisible novelties inasmuch as it alters the nature of the preliminary reference procedure by modifying the place and role within that procedure of both the parties to the main proceedings and the national courts of last instance. After recalling the facts and context of the case and present the Advocate General’s Opinion together with the main findings of the Court, I will offer an analysis of the judgment with a focus on those visible and invisible novelties.

The facts and context of the case

The applicants in the main proceedings (a temporary association of undertakings providing various services) and the defendant (the Italian railway infrastructure manager) concluded a public contract for, inter alia, the supply of cleaning services for national railway infrastructure. During the performance of that contract, the applicants asked the defendant to review the contract price. The defendant refused. The applicants challenged that refusal before an Italian regional court and, subsequently, before the Consiglio di Stato (the Italian supreme administrative court) on the grounds that Italian law, which allowed the exclusion of price review, was in breach of EU law.

That case gave rise to two preliminary references to the Court of Justice: Case C-152/17 and, subsequently, Case C-569/19.

In Case C-152/17, the Consiglio di Stato asked the Court to interpret a first set of provisions of EU law (especially Article 56 TFEU, Article 16 of the Charter of Fundamental Rights of the European Union (the Charter) and Directive 2004/17 on public contracts in the special sectors) ³ and to assess the validity of the latter directive. In its judgment of 19 April 2018, ⁴ a three-judge chamber found inadmissible most of the request for a preliminary ruling for the non-applicability or lack of relevance of the EU law provisions raised by the referring court. The Court only offered an interpretation of Directive 2004/17 to the effect that the latter did not preclude national rules which do not provide for price review after a contract has been awarded in the sectors covered by that directive.

Six months after the delivery of the Court of Justice’s judgment, the referring court held a public hearing where the applicants in the main proceedings asked it to refer further questions for a preliminary ruling in relation to other substantive provisions of EU law that had not been brought up within the first request. The Consiglio di Stato thus referred the case for a second time to the Court of Justice. That is Case C-561/19, which is the subject of this commentary.

The novelty of Case C-561/19 compared to Case C-152/17 lies in the fact that, beyond substantive questions in the matter of public procurement that will not be examined here, ⁵ the Consiglio di Stato also referred a question regarding the scope of Article 267, paragraph 3, TFEU in the specific circumstances of the case: is a national court of last instance required to make a reference on a question of interpretation of EU law even where that question has been submitted to it by a party to the proceedings not in its initial pleading but subsequently, in particular after the case has been set down for judgment for the first time or even after a prior preliminary reference in the same case? It is in the context of that latter question that Case C-561/19 has been considered as raising seminal constitutional issues regarding the duty to refer and, more broadly, the respective roles and places within the preliminary reference procedure of the Court of Justice, of national courts of last instance and of the parties to the main proceedings.

The Opinion of Advocate General Bobek

In his Opinion, ⁶ Advocate General Michal Bobek acknowledged from the outset that the question regarding Article 267 TFEU posed by the referring court could be answered without further ado, since the Court could content itself with restating its classic case law on the duty to refer and concluding that the referring court may bring the matter again to the Court. ⁷ However, Advocate General Bobek conspicuously took the view that the case at hand was a good case not only to clarify but also to revisit Cilfit. After explaining, in a 39-page, quite ‘academic’ Opinion, why the case could be easily dealt with, he painstakingly set the stage by teasing out the various problems associated with Cilfit and the exceptions to the duty to refer before dedicating the last third of his Opinion to proposing a relatively new approach to the duty to refer.

Setting the stage

Advocate General Bobek started with recalling the functions of the preliminary reference procedure. While the micro-purpose of the preliminary reference procedure in general is to assist a national court in settling a specific dispute, the macro-function is to secure uniform interpretation of EU law. That macro-function is key to understanding the fact that national courts of last instance have a duty to refer since it is at that level within the member states that divergences within national case law can be ultimately corrected and that the rights conferred upon individuals by EU law may be vindicated. ⁸

After presenting the judgment in Cilfit and the three exceptions to the duty to refer, Advocate General Bobek focused on the third one and set out at length the various shortcomings associated with the ‘Cilfit criteria’ regarding the identification of an acte clair/reasonable interpretive doubt. ⁹ First, he noted that the Court of Justice itself was not fully consistent in its own application of those criteria. ¹⁰ In the light of the judgments in X and van Dijk ¹¹ and Ferreira da Silva e Brito, ¹² he concluded that the Court was seemingly more concerned by interpretive divergences across the Union rather than within a single member state, where the Court appears more relaxed. ¹³ Second, after underlining that the acte clair doctrine was plugged into EU law from a French theory which had a different purpose, ¹⁴ Advocate General Bobek was critical of the fact that, according to him, the assessment of the existence of a duty to refer had shifted over time from the requirement of uniform interpretation to that of individual application. ¹⁵ Third, in the wake of previous advocates general, Advocate General Bobek noted that the Cilfit criteria are actually not applied by national courts of last instance. He gave several illustrations where the latter have substituted them with their own criteria. ¹⁶

Finally, Bobek also regretted the asymmetry between the uncertainty regarding the assessment of acte clair (and the varying practices at the national and European levels) and its enforcement. Some national constitutional courts check compliance with the duty to refer as part of their evaluation of the respect for the relevant right to a lawful judge that their constitution provides for. ¹⁷ By the same token, the European Court of Human Rights examines on the basis of Article 6(1) of the Convention whether national courts of last instance duly justify their decisions not to refer in the light of the three exceptions to the duty to refer laid down in Cilfit. ¹⁸ However, it is with the Court of Justice itself, in particular the judgment in Commission v France, ¹⁹ that Advocate General Bobek was more critical vis-à-vis the recent ease with which infringement proceedings can be successfully launched against a member state for the failure of its last instance national courts to seize the Court, ²⁰ especially at a time where the Court seemed to have somehow relaxed its assessment of acte clair in X and van Dijk.

A new three-pronged test to determine whether a national court of last instance is to refer questions to the Court

In the last section of his Opinion, Advocate General Bobek made a pragmatic proposal to revisit the duty to refer so that the scope of the latter would match its macro-function: preventing judicial divergences across and within the member states, in other words ensuring the uniform interpretation of EU legal rules as opposed to the uniformity of individual outcomes. ²¹ That proposal is an explicit plea in favour of realism and mutual trust: realism regarding both the (in)ability or reluctance of national courts of last instance to apply the Cilfit criteria and the type and degree of uniformity that should be aimed at the EU level; ²² trust towards national courts of last instance, which are increasingly compliant with the duty to refer. ²³ That proposal, which arguably does not depart from the Cilfit criteria, in particular acte clair, but rather tries to streamline their application and make their outcome more predictable, consists of the new three-pronged test:

a court or a tribunal of a Member State against whose decisions there is no judicial remedy under national law is to refer the case to the Court of Justice provided that it raises (i) a general issue of interpretation of EU law (as opposed to its application); (ii) to which there is objectively more than one reasonably possible interpretation; (iii) for which the answer cannot be inferred from the existing case-law of the Court of Justice (or with regard to which the referring court wishes to depart). ²⁴

As regards the first condition, ‘the duty to refer ought to be triggered whenever a national court of last instance is confronted with an issue of interpretation of EU law, formulated at a reasonable and appropriate level of abstraction’ ²⁵ and ‘the question of interpretation ought to concern a general, potentially recurring question of interpretation of EU law’. ²⁶

As far as the second condition is concerned, ‘where there are two or more potential interpretations proposed before the national court of last instance, the duty to refer becomes strict’. ²⁷ That condition is the one that Advocate General Bobek developed most as a way to polish acte clair:

Provided that there is indeed a divergence in possible legal interpretation, demonstrated by possible alternatives, then, using the CILFIT terminology, (reasonable) doubt can be considered to have been objectively and externally established in the dispute before them and the duty to refer in the interest of ensuring uniform interpretation of EU law cannot then be ignored. ²⁸

Such divergence would not necessarily have to be established by national courts themselves, but the latter ‘should recognise that there is objectively a divergence in legal interpretation if that divergence has been brought expressly to their attention by any of the actors in the proceedings before them, in particular by the parties themselves’. ²⁹

As regards the third condition, Advocate General Bobek refined Da Costa and the exception to the duty to refer relating to the existence of established case law that may be relied on in a similar case. ‘Established case law’ should indeed be understood as referring to general interpretations of EU law rules, whether it be several judgments confirming a certain line of case law or merely a single prior judgment when that latter appears sufficiently authoritative.

Last but not least, Advocate General Bobek proposed, on the basis of Article 47 of the Charter to impose on national courts of last instance a new, ‘correlating duty to specifically and adequately state reasons’ ³⁰ for not referring questions of EU law to the Court:

should a national court of last instance be of the view that, even if faced with an issue of interpretation of EU law in the main proceedings, one of the three conditions is not met, that court is obliged to identify clearly which one of the three conditions is not met and state the reasons why it believes that to be the case. ³¹

The judgment

In its judgment, the Court started by restating general ideas drawn from previous case law on the objectives of the preliminary reference procedure, notably the fact that it aims at ‘securing uniform interpretation of EU law’ ³² and ‘to ensure that, in all circumstances, EU law has the same effect in all Member States and thus to avoid divergences in its interpretation’. ³³ The Court then moved on to the duty to refer of national courts of last instance. It repeated in an (almost) unaltered manner, and on the basis of its earlier case law the three classic exceptions to the duty to refer: (a) when the question on EU law that is raised is irrelevant; (b) where the EU law provision in question has already been interpreted by the Court; (c) where the correct interpretation of EU law is so obvious as to leave no scope for any reasonable doubt. ³⁴

Regarding that third condition, the Court also restated in full the various elements that are to be looked at to determine the existence of reasonable interpretive doubt. ³⁵ However, it brought in a few clarifications regarding that evaluation. First, re comparison between the language versions: the Court makes it clear that national courts of last instance are not required to examine each of the language versions, but ‘must bear in mind those divergences between the various language versions of that provision of which it is aware, in particular when those divergences are set out by the parties and are verified’. ³⁶ Second, it makes clear that the need to avoid interpretive divergences applies both to divergences ‘among the courts of a Member State or between the courts of different Member States’. ³⁷ Third, the ‘national court or tribunal of last instance must be convinced that the matter would be equally obvious to the other courts or tribunals of last instance of the Member States and to the Court of Justice’. ³⁸ Fourth, the Court held that ‘the mere fact that a provision of EU law may be interpreted in another way or several other ways, in so far as none of them seem sufficiently plausible for the national court or tribunal concerned…, is not sufficient for the view to be taken that there is reasonable doubt as to the correct interpretation of that provision’. ³⁹ Fifth, it is now the interpretation of EU law (rather than its application) that must leave no scope for reasonable doubt. ⁴⁰

Last but not least, the Court of Justice imposed on national courts of last instance a duty to state reasons when they decide to refrain from referring: ‘it follows from the system established by Article 267 TFEU, read in the light of the second paragraph of Article 47 of the Charter’, that a national court of last instance that is of the view that it does not have to refer because the situation at hand falls within one of the three exceptions to the duty to refer must, however, state the reasons for its decision not to refer. ⁴¹

The Court eventually replied to the specific question on Article 267 TFEU that was posed to it by the referring court regarding the potential relevance of the fact that the Court had previously answered the referring court in the same set of proceedings. ⁴² After recalling that Article 267 TFEU is not a means of redress for the parties but is a procedure at the discretion of national courts, ⁴³ the Court restated the latter’s duty to refer if none of the three abovementioned exceptions is applicable. ⁴⁴ In particular, if the Court’s answer still proves necessary, it must seize the Court even after to a prior preliminary judgment. ⁴⁵ However, a member state may subject the admissibility of new questions to certain national rules as long as the latter respect the principles of equivalence and effectiveness. ⁴⁶

Analysis

Has Cilfit become the inescapable horizon for national courts of last instance to determine whether they are obliged to refer to the Court of Justice questions on the interpretation of EU law under Article 267, third paragraph, TFEU? It would appear so in the light of the judgment at hand, where the Court not only referred to Cilfit 10 times but also restated in full the three exceptions to the duty to refer that were laid down therein. The continuity with Cilfit, in particular the contested acte clair doctrine, ⁴⁷ is what strikes at first read. Cilfit is still good law. Cilfit still fits. At the same time, however, that continuity cannot conceal a few visible novelties as regards the determination of what is an ‘acte clair’ and – most significantly – the new duty imposed on national courts of last instance to state reasons. A closer look even conjures up the invisible, systemic – and perhaps unintended – consequences of the judgment on the nature of the preliminary reference procedure and the place and role of both the parties to the main proceedings and national courts of last instance with regard to EU law-based arguments.

Since the continuity with Cilfit is what is palpable in the first place in this judgment, I will start by setting out the likely reasons why the Court has preferred to stick to Cilfit together with the three classic exceptions to the duty to refer laid down therein. I will then move on to the visible novelties of the judgment in Consorzio Italian Management and ultimately conclude with the invisible novelties and their impact on the overall architecture of the preliminary reference procedure.

Why prefer the (relative) status quo regarding the scope of the duty to refer?

Consorzio Italian Management was one of those ‘good cases’ with which to overhaul the established case law, in casu Cilfit, for at least two reasons. First, cases where a national court expressly asks questions on the scope of the duty to refer are relatively scarce before the Court of Justice. Second, because that case did not raise substantive issues of EU law, the Court could afford to focus on more procedural issues, such as the scope of the duty to refer. Yet, the Court did not seize that opportunity to significantly revisit the exceptions to the duty to refer. Instead, the Court opted to more or less maintain the status quo, by restating the Cilfit exceptions in (almost) the exact same words so that the latter now appear durably embedded in EU law. Several reasons may be adduced to explain why the Court has opted for a marginal adjustment of the Cilfit exceptions.

First, there are reasons external to the Court, that is the judicial and political contexts of the current times where the primacy of EU law and the preliminary reference procedure are challenged by some constitutional courts. The timing has admittedly become less suitable to reconsider the Cilfit exceptions for fear of disrupting the whole system established by Article 267 TFEU. ⁴⁸ Cilfit might be a sleeping dog that should not be disturbed within the Court ⁴⁹ but the danger rather comes from other already barking (and biting) organs in the Member States which either limit referrals ⁵⁰ or disregard preliminary judgments. ⁵¹ Arguably, it is in response to such situations – especially the PSPP ‘precedent’ of the German Federal Constitutional Court – that the Court of Justice has restated the duty to refer in paragraph 38 of its judgment: ‘a national court or tribunal of last instance must make such reference when it encounters difficulties in understanding the scope of the judgment of the Court’. ⁵²

Second, there are reasons that are inherent in judicial politics. It is common knowledge that overruling (or call it overhauling) is generally not an easy thing to undertake. Overruling may generate legal uncertainty, political turmoil and reputational costs for courts of law, as recently illustrated by several controversial judgments of the United States Supreme Court. Therefore, thorough reconsideration of previous case law should be confined to cases where a court of law is satisfied that its new stance is for the better, whatever ‘better’ may mean in that context. Arguably, when there is no clear reason why a new solution and reasoning should be devised, it is preferable for the bench to refrain from shaking up established case law and being innovative just for the sake of it, especially when that case law has acquired some kind of precedential value with the passage of time. Rather like in economics, where some companies are ‘too big to fail’, in law some judgments may be ‘too established to be overruled’. In that respect, Cilfit has become so engrained in EU law that a deep revisiting could certainly disrupt the preliminary reference procedure as the keystone of EU law.

Third, there are reasons internal to Cilfit itself that explain the latter’s confirmation by the Court in Consorzio Italian Management: having stood the test of time, Cilfit is perhaps not that bad after all. On the one hand, its ‘flexibility’ (to be understood as ‘variability’) is certainly at odds with both legal certainty and uniformity, somewhat paradoxically for a tool that was designed to ensure uniform interpretation. On the other hand, however, that flexibility may also be seen as a strength that generates positive externalities. ⁵³ ‘Eppur si muove!’ Indeed, all national courts know about Cilfit and they perhaps like it as much as EU law students do, ⁵⁴ precisely for that relative flexibility that it offers as regards the duty to refer. Although national courts tend to use different standards, ⁵⁵ there is overall no shortage of preliminary references made by national courts. ⁵⁶

It follows from those various reasons that the Court of Justice should not be overly criticised for restating Cilfit and its set of exceptions, especially in view of the novelties that were brought in.

Some minor clarifications regarding the assessment of the existence of reasonable interpretive doubt

As explained above in the subsection summarising the main findings of the judgment, the first set of novelties that the Court has introduced in its judgment concern Cilfit’s third exception to the duty to refer. The Court indeed provided further – albeit limited ⁵⁷ – guidance as to the evaluation of acte clair/reasonable interpretive doubt by making some clarifications.

As a preliminary remark, since the Court does not commonly engage with the arguments of its Advocates General, ⁵⁸ it is worth noting that all clarifications have been made in response to the specific concerns that were raised by Advocate General Bobek with regard to the uncertainty surrounding the reasonable doubt exception. Although the Court of Justice did not wish to embrace the Advocate General’s test as a whole, the Court still met the latter’s concerns individually.

Clarifications are, of course, always welcome when they enhance legal certainly. When looking in detail at each of the five clarifications brought in by Consorzio Italian Management, they certainly facilitate the assessment of the existence of reasonable interpretive doubt. However, most of them generate their own interpretive issues and do not make the outcome of that assessment significantly more predictable.

First, on the comparative evaluation of language versions, the Court does not explain in detail exactly what is expected from national courts. That should not, however, be seen as a major problem since it can be inferred from the judgment that the Court is fully aware of how daunting that task can be. It can be assumed that looking at one or two foreign versions should be sufficient. There should thus no longer be any fear that national courts of last instance have to ‘turn into comparative EU law research centres, whereby they … ex officio carry out searches into the case-law of other national courts in other Member States or start looking proactively into finding interpretative problems’. ⁵⁹ In any event, since the text of a provision is not conclusive of its actual meaning (the latter depending also on its context and purpose), neither is the fact that the tenor of an EU law provision is the same across the language versions decisive as to (the absence of) reasonable doubt. It is only where the language versions immediately open up different interpretive avenues that the existence of reasonable doubt should be presumed.

Second, regarding the source(s) of interpretive divergences that are relevant for the purposes of the assessment of reasonable doubt, two aspects of the judgment are worth developing: (a) the exclusive focus on judicial interpretation; (b) the uncertainty regarding the relevance of lower courts’ interpretive doubt.

There is no mention of divergences that would arise from administrative interpretations of EU law. It thus appears that the Court only considers judicial interpretation and seems to disregard administrative interpretations, in the wake of Intermodal Transports. ⁶⁰ On the one hand, such a stance is understandable because what ultimately matters in court-centred liberal democracies is judicial interpretation. That said, assessing the existence of reasonable interpretive doubt for the purposes of determining whether there is a duty to refer is a different task from ascertaining the meaning of a legal provision in order for it to be judicially enforced in a specific case. As a consequence, the assumption that reasonable doubt can only arise from judicial interpretive divergences is questionable. Such doubt is, for instance, likely to arise from a ministerial circular (especially in tax matters) that usually interprets the relevant legal rules in a certain area. In that type of case, it would be somewhat puzzling if a court before which a question of interpretation of EU law is raised could refrain from bringing the matter before the Court ‘just’ because the competing interpretation does not originate from a fellow judge but from the government. It would thus seem commendable that national courts of last instance also take divergences originating in an administrative body as an indicator of the existence of reasonable interpretive doubt, which Consorzio Italian Management does not prevent them from doing anyway.

In addition, it is uncertain whether lower courts’ interpretive doubt is relevant for a national court of last instance to determine whether an EU law provision is clear. On the one hand, the Court has altered Cilfit by specifying that the interpretation in issue should be equally obvious to the Court and to the national courts of last instance. ⁶¹ On the other hand, the Court has not been specific when stating that national courts of last instance should be vigilant when there are divergences ‘among the courts of a Member State or between the courts of different Member States’. ⁶² Arguably, since the Court relied on both X and van Dijk and Ferreira da Silva e Brito in its judgment, it is sensible to take the view that interpretive doubt arising from lower courts does not matter between member states. As far as domestic courts are concerned, such an approach is more debatable, though.

Third, regarding the consequences to be drawn from the existence of several plausible interpretations, the Court of Justice’s point is not entirely clear. In the somewhat convoluted paragraph 48 of the judgment, the Court has specifically replied to the second prong of Advocate General Bobek’s proposed test regarding the existence of ‘more than one reasonably possible’ interpretation. It is submitted here that the Court has actually embraced the proposal made by the Advocate General on that matter. The Court indeed seems to distinguish between possible and plausible interpretations. Whereas a range of simply possible interpretive options does not create reasonable doubt in the mind of a judge, the existence of several plausible ones (those that are thus ‘reasonably possible’) does, leading in turn to a duty to refer. Such distinction is perhaps not a paragon of clarity since, arguably, an impossible interpretation is no interpretation. However, in order to refine the evaluation of acte clair, it is certainly appropriate to focus on those interpretations only that are plausible, i.e. not far-fetched, in the light of the text (in some of its language versions), the context and purpose of the EU law provision in issue.

Fourth and finally, Consorzio Italian Management has restated its three Cilfit exceptions to the duty to refer in the very same words but one. In respect of acte clair, it is no longer the correct application, but the correct interpretation of EU law that should leave no scope for any reasonable doubt. Such change appears to ‘cautious[ly]’ ⁶³ take into account Advocate General Bobek’s criticism vis-à-vis the drafting of requests for a preliminary ruling towards the application of EU law by the Court instead of the general and abstract interpretation thereof. Although it is admittedly difficult to clearly determine what belong respectively to the interpretation and to the application realms, this is no reason for not trying to distinguish between them in accordance with the classic – and meaningful – distribution of tasks between the Court (interpretation of EU law) and the national courts (application of EU law). ⁶⁴ It is now for the Court to show that that change is not purely nominal by altering its actual practice to the effect of providing fewer fact-tailored answers.

The new duty to reasons: a major novelty with constitutional implications

While Consorzio Italian Management was, for the reasons mentioned above, not considered to be a good case in which to rethink the exceptions to the duty to refer, it was at least deemed good enough to introduce a major novelty in the wake of the Advocate General’s Opinion, namely the duty to state reasons when a national court of last instance decides not to refer questions on interpretation of EU law to the Court of Justice. ⁶⁵ In its judgment, the Court derived such duty from a combined reading of Article 267 TFEU and Article 47 of the Charter of Fundamental Rights. Within that duty, national courts of last instance:

must show either that the question of EU law raised is irrelevant for the resolution of the dispute, or that the interpretation of the EU law provision concerned is based on the Court’s case-law or, in the absence of such case-law, that the interpretation of EU law was so obvious to the national court or tribunal of last instance as to leave no scope for any reasonable doubt. ⁶⁶

In other words, national courts of last instance must establish that they fall within one of the three Cilfit exceptions to the duty to refer in order to ground their decisions not to refer.

With the introduction of the duty to state reasons, the Court has made a welcome move in at least three respects: first, it promotes convergence between the Court of Justice of the European Union and the European Court of Human Rights, since the latter is regularly called on to review national judgments in the light of the right to a fair hearing under Article 6 of the Convention on Human Rights when applicants argue that national courts of last instance have infringed that right by refusing to seize the Court under Article 267 TFEU; second, it has found a clever way to enhance the effectiveness of the duty to refer under EU law by the proceduralisation of the latter; third, and in relation to the previous point, the Court facilitates but also necessarily redirects the enforcement of the duty to refer by shifting the focus from the duty to refer to the duty to state reasons.

After presenting the origins of the duty to state reasons in the case law of the European Court of Human Rights, I will set out the issues regarding the exact contours of the duty to state reasons, and the evolution of enforcement mechanisms that that duty may bring with it.

(i) Taking a cue from the European Court of Human Rights

Although the Court of Justice did not cite the case law of the European Court of Human Rights in its judgment, it would be hard to deny the latter’s influence on the former as far as that new duty to state reasons is concerned. Indeed, that duty was originally devised by the European Court of Human Rights within those situations in which it had to examine refusals of national courts to refer a case to the Court of Justice, both in relation to the application of the Bosphorus presumption and, autonomously, with regard to Article 6(1) of the Convention.

Regarding the first situation, it is well-known that ever since Bosphorus v Ireland, ⁶⁷ the European Court of Human Rights does not usually look at how member states apply EU law, when such an examination would be tantamount to looking at EU compliance with the Convention, i.e. where member states enjoy no discretion in implementing EU law obligations. Although the European Court of Human Rights has accepted that it will leave that task to the Court of Justice, it is on the condition, inter alia, that the deployment of the full potential of the supervisory mechanism provided for by EU law is ensured by national courts of last instance. ⁶⁸ The European Court of Human Rights does not hesitate to enforce that condition. ⁶⁹

It is, however, in the second situation that the European Court of Human Rights has specifically derived a duty to state reasons for national courts of last instance from Article 6(1) of the Convention. In Ullens de Schooten v Belgium, ⁷⁰ the European Court of Human Rights refused to recognise on the basis of the Convention any ‘right to have a case referred by a domestic court to another national or international authority for a preliminary ruling’. ⁷¹ However, it took the view that national courts of last instance which refuse to refer to the Court of Justice a preliminary question on the interpretation of EU law that has been raised before them are obliged to give reasons for their refusal in the light of the three exceptions provided for in the Cilfit case law of the Court of Justice. ⁷² In coming up with a similar duty to state reasons, the Court of Justice has thus embraced the approach of the European Court of Human Rights.

(ii) The contours of the duty to state reasons

Now that the Court of Justice has durably sealed Cilfit and its exceptions to the duty to refer, it is only logical that the duty to state reasons be tailored to those exceptions. National courts of last instance should explain in concreto which one of those three situations applies to the case in hand to justify their decision not to refer.

As regards the first exception, if the EU law question raised is irrelevant for the purposes of settling a case, stating reasons should be relatively straightforward. The national court will have to explain why that provision is inapplicable in the factual circumstances of a case so that it is obviously not necessary to ascertain its true meaning. ⁷³

By contrast, should the question appear potentially relevant, national courts of last instance must determine whether the second or the third exceptions come into play. They will then be under a heightened level of scrutiny since they will have to explain why they can interpret EU law themselves without endangering the latter’s uniform interpretation. In that respect, national courts must either make argument based on the most relevant case law of the Court, or demonstrate that the EU law provision in issue is clear. Each of those situations raises its own set of difficulties.

Within the first scenario (corresponding to acte éclairé), exactly what kind of argument may be made? An easy case is the one where a national court of last instance takes the view that a previous judgment (or jurisprudential line) applies, since it should then cite and engage with that case law in order to duly explain why that case law applies to the case at hand and what solution derives from it. A more uncertain case is the one where a national court of last instance is fully aware of the Court’s case law but is of the view that it is not appliable to the factual situation at hand. In an ideal world, such a ‘distinguishing’ technique, which is the result of the (negative) application of EU law to the national context, should be allowed since it does not challenge the interpretation of EU law. In practice, however, the national court will have to provide a detailed statement of reasons in such a scenario to prove its point. Admittedly, the line may be thin between ‘true’ and ‘false’ departures from previous case law, that is failing to apply a previous judgment either because the national court simply does not want to (contestation of the Court’s judgment, thus true departure) or because it considers it too different to be relevant (distinguishing, thus false departure). Unlike Advocate General Bobek, ⁷⁴ however, the Court did not expressly impose a requirement to comply with the duty to refer when a national court of last instance departs from the Court’s case law, which leaves the issue open. It is, however, to be expected, especially in the light of the judgment in Commission v France, ⁷⁵ that the Court will prove demanding vis-à-vis national courts of last instance so that the latter sufficiently explain: (a) why the situation at hand is different from a situation dealt with in a previous judgment; and (b) how EU law is then irrelevant (first exception to the duty to refer), or why EU law is clear as far as the situation at hand is concerned (third exception to the duty to refer). It follows that a national court of last instance will thus fall back on the two other exceptions to the duty to refer and cannot simply rely on the Court’s case law to dismiss it.

Within the second scenario (corresponding to acte clair), it will not be straightforward in practice for a national court of last instance to show that the interpretation of the EU law provision in issue ‘does not leave scope for any reasonable doubt’. Although such task will undoubtedly be easier to carry out thanks to the clarifications that were brought in by the Court, it has been explained above how those clarifications generate their own uncertainties. More generally, as it stands, the acte clair exception to the duty to refer still expects too much from those ‘mortal national judges not possessing the qualities, time, and resources of Dworkin’s Judge Hercules’, ⁷⁶ namely to think just like the Court of Justice and apply the latter’s methods of interpretation consisting in looking at the text (in the light of other language versions), context and purpose of the provision in issue. On the one hand, national courts should certainly be more acquainted with those methods nowadays. On the other hand, expecting them, for the mere purpose of deciding whether they must refer a case to the Court, to interpret EU law in the way the Court does is paradoxical, since it ultimately suggests that the latter’s interpretive guidance would not be needed.

Be that as it may, because the acte clair exception remains blurred in its contours, the statement of reasons in that regard cannot be overly strict. Arguably, the expected level of detail of the statement of reasons will vary from one situation to another, depending on the perceived ‘threat’ to the authority of EU law. Yet again, the European Court of Human Rights standard, which shall be a minimum under EU law in accordance with Article 52(3) of the Charter, ⁷⁷ provides further guidance as to what should be the intensity of the review of that statement of reasons. The European Court of Human Rights has indeed proved rather reasonable in its assessment of the duty to state reasons. Hence, a summary reasoning suffices when a case raises ‘no fundamentally important legal issues or had no prospects of success’. ⁷⁸ Further, ‘in concreto, the reasons for the rejection of the request for a preliminary ruling under the Cilfit criteria can be deduced from the reasoning of the remainder of the decision given by the court in question (or from reasons considered implicit in the decision rejecting the request)’. ⁷⁹ In more difficult or important cases, the European Court of Human Rights expects national courts of last instance to explicitly refer to the three Cilfit criteria and explain which of those criteria was used as the basis for deciding not to transmit the case to the Court of Justice. ⁸⁰ Clearly, those are minimum requirements. The Court of Justice will probably have somewhat higher expectations, including when it comes to acte clair.

(iii) Enforcement issues

As already demonstrated by the case law of the European Court of Human Rights, the duty to state reasons necessarily bears consequences at the enforcement level. In order to establish that they can escape their duty to refer, national courts of last instance must now fulfil another duty: the duty to explain why and how they decided to keep the case for themselves and not share it with the Court of Justice. That requires evidence as to the lack, insufficiency or distortion of the statement of reasons put forward by national courts of last instance when they decide not to refer issues of EU law that were raised by the parties to the main proceedings. As a consequence, a national court of last instance should not be sanctioned for non-compliance with the duty to refer itself but for non-compliance with the duty to state reasons.

With the focus thereby shifting from the duty to refer to the duty to state reasons, Consorzio Italian Management paves the ground for a necessary redirection in terms of remedies when it comes to the sanctioning of violations of Article 267, third paragraph, TFEU: member state liability now appears more suitable than infringement proceedings to enforce the duty to refer.

Until now, despite Köbler ⁸¹ opening up that possibility, it was virtually impossible to trigger member state liability purely on the basis of a failure to comply with the duty to refer in view of the highly relative nature of that obligation and the fact that there is no subjective right to be granted a request for a preliminary reference. ⁸² That situation has changed with the new duty to state reasons. The latter arguably makes it easier for the parties to the proceedings to obtain damages. ⁸³ Subject to confirmation in future cases, a subjective right for the parties to the main proceedings to obtain a statement of reasons can be derived from the duty to state reasons. Those parties could then rely, before the competent national court of law, on the fact that the right has been violated by a national court of last instance by a faulty statement of reasons for not referring a case to the Court of Justice. ⁸⁴ As a consequence, they could possibly claim that they suffered loss of opportunity or moral damage therefrom. ⁸⁵

Conversely, the shift from the duty to refer to the duty to state reasons renders infringement proceedings less likely. With the new duty to state reasons, it is primarily the failure to comply with that duty that could give rise to findings of infringement. It is assumed that infringement proceedings will be launched only against national courts of last instance in the exceptional circumstances of either a repetitive, systemic lack of statement of reasons by a national court of last instance, or a gross and intentional misrepresentation of the situation at hand by a national court of last instance made up to eschew its duty to refer – in other words, an egregious breach of the duty to refer that cannot find any justification.

Such limitation of the scope of infringement proceedings appears appropriate in view of the specific nature of the judicial function, the discretion left to national courts of last instance by Cilfit and the ensuing uncertainty inherent in the scope of the duty to refer. As stated above, Cilfit should still be seen as good case law only to the extent that the Court itself fully acknowledges that flexibility and accept the consequences thereof by deciding not to enforce the duty to refer as such. In Commission v France, ⁸⁶ the Court of Justice had proved quite harsh on national courts of last instance in castigating the French Conseil d’Etat for a violation of a duty to refer which had never been enforced by the Court of Justice in the past. ⁸⁷ Retaining the loose Cilfit because of its flexibility is one thing; it is quite another to enforce it in a strict manner against unique judicial partners, especially within proceedings that also have a name-and-shame dimension. What should primarily be enforced now is instead the duty to state reasons.

Conclusion: systemic consequences on the preliminary reference procedure

The novelties introduced by the judgment at hand go further than those that are obvious and visible; they also include some largely invisible points. These take the form of those – probably unintended – consequences of the judgment on the nature of the preliminary reference procedure itself as regards the relationship between the Court of Justice and national courts of last instance and the (enhanced) role of the parties to the main proceedings.

First, with the introduction of the duty to state reasons, Consorzio Italian Management places national courts of last instance under the monitoring of various actors, in the first place the Commission and the Court of Justice, but also the parties themselves and lower national courts. Certainly, that supervision, which is not entirely new, should be put into perspective. National courts of last instance were already under the supervision of the European Court of Human Rights as regards the duty to refer and some of them were also under the scrutiny of their constitutional courts. ⁸⁸ In addition, with Commission v France, national courts of last instance were also monitored by the Commission and, ultimately, by the Court of Justice. Despite the unlikelihood in the future of a repetition of a judgment such as Commission v France (at any rate in its current form), that new type of supervision requires national courts of last instance to drastically change their practices by systematically engaging with EU law arguments instead of dismissing them without further ado. That admittedly alters their position within the preliminary reference procedure, almost making them look like parties who must defend themselves by giving reasons at all times when an EU law issue is raised. At the same time, that position should have arguably been theirs from the very beginning of the European Communities, since it goes hand in hand with the duty to refer. Until now, national courts of last instance had somehow acquired bad habits in thinking that their duty to refer was ultimately optional, having discretion to refer like any other national court. With Consorzio Italian Management, national courts of last instance keep some discretion but that discretion is rightly curtailed by their duty to state reasons.

Second, Consorzio Italian Management significantly enhances the place of the parties to the main proceedings within the preliminary reference procedure. That statement may at first sound odd because of their traditionally limited role and the absence of a subjective right for the parties to the main proceedings, be it under EU law or ECHR law, to a preliminary reference. Ever since Cilfit, preliminary references do ‘not constitute a means of redress available to the parties to a case pending before a national court or tribunal’ ⁸⁹ and are ‘completely independent of any initiative by the parties’. ⁹⁰ Recognising such a right ⁹¹ would have more drawbacks than advantages. Above all, it would entirely change the nature of the preliminary reference procedure and the respective roles of the Court of Justice and national courts of last instance.

Despite the absence of such right for now, the role of private parties ends up strengthened in relation to the assessment of the existence of reasonable doubt and, inherently, within the duty to state reasons. As regards the former, the judgment contains various references to the parties that clearly suggest that, when deciding whether to refer or not, national courts of last instance are certainly not bound by the parties’ arguments but should take them seriously and engage with them to come to its decision. For instance, a national court of last instance ‘must bear in mind those divergences between the various language versions of that provision of which it is aware, in particular when those divergences are set out by the parties and are verified’. ⁹² Further, where a court of last instance ‘is made aware of the existence of diverging lines of case-law’, it ‘must be particularly vigilant in its assessment of whether or not there any reasonable doubt’. ⁹³

As regards the duty to state reasons, this does not purely fulfil the function of ensuring the effectiveness of the duty to refer. The duty of public authorities at large to state reasons, as a corollary of an individual’s right to defence, is classically associated with private parties. That duty thus introduces a strong element of subjectivity into the otherwise objective nature of the preliminary reference procedure. In the absence of an enforceable right for the parties to obtain a preliminary reference, there is arguably now an enforceable right under EU law to receive a statement of reasons in reply to their specific arguments. In so doing, although it primarily remains a cooperation mechanism before courts, the preliminary reference procedure may unexpectedly be undergoing a reorientation towards the parties. That is probably yet another unintended consequence of a minimalist judgment with a far-reaching constitutional impact.