The Unfolding Story of Judicial Dialogue in the EU: The Coercive and Persuasive Motives Behind the Participation of Belgian Highest Courts in the Preliminary Ruling Procedure

Abstract

The story of national judges’ participation in the process of dialogue with the Court of Justice of the European Union (ECJ) pursuant to Article 267 TFEU is a well-known one. There have been many scholarly efforts to explain and theorize the—lack of—participation of national courts in the mechanism. Indeed, various constitutional, institutional, procedural, cultural, political, and individual factors are pointed out to explain this engagement. Yet, it may come as a surprise that still much remains unknown. The available empirical data concerning the potential causes of the national judges’—non—participation is limited to a number of EU member states and is mostly confined to lower courts judges. It is the aim of this Article to partly fill the void in the literature and to provide novel empirical data gained in the course of semi-structured interviews among judges and court référendaires of three Belgian highest courts. The interviews provide insight into the possible factors driving concerned national judges when deciding (not) to refer preliminary questions, including the role and significance of the obligation to refer which is to be found in Article 267 TFEU and the Cilfit-criteria.

A. Introduction

The story of national judges’ participation in the process of European integration, or more precisely, in the process of dialogue with the Court of Justice of the European Union (ECJ or Court) pursuant to Article 267 TFEU, is a well-known one. This system of dialogue in which national courts can refer questions on interpretation or validity of EU law to the Court is widely considered as having been essential for the construction and development of EU’s legal order. ¹ The Court itself considers the procedure to constitute the “keystone” of EU’s judicial system. ² The functioning of the preliminary ruling procedure, however, very much depends on the willingness of national judges to resort to the mechanism. ³ Many academics in the fields of political science, law, and public administration have been preoccupied with this issue. ⁴ They have worked tirelessly to develop theoretical and empirical explanations for how judges participate in the mechanism, as well as potential causes and motivations for judges’ (non) participation. Indeed, various constitutional, institutional, procedural, cultural, political, and individual factors are pointed out to explain this engagement. ⁵ Yet, it may come as a surprise that still much remains unknown. The empirical data concerning the potential causes of the national judges’ (non) participation that has so far been made available is limited to a small number of (former) EU member states, including Croatia, Germany, the Netherlands, Ireland, Italy, Poland, Slovenia, Spain, Sweden, and the UK. ⁶ Insights into national judiciaries of the remaining EU Member States is however vital for gaining a fuller picture of the functioning of the preliminary reference mechanism, as it is generally assumed that data from one Member State cannot be generalized and used to explain judges’ behavior in another EU member state. ⁷

It is the very aim of this Article to partly fill the void in the literature and to investigate the various reasons behind (non) participation of Belgian highest court judges in the procedure. The ultimate objective of this Article is to advance the field’s theoretical understanding of judicial behavior in the framework of the preliminary ruling procedure by reexamining and advancing the “grand theories” already in existence. For that purpose, we provide novel empirical data gained from semi-structured interviews among judges and court référendaires—better known as legal secretaries—of three Belgian highest courts. It needs to be underscored that the empirical data that has been made available so far is in its vast majority confined to lower court judges. However, an examination of reasons (not) to refer by highest national courts is particularly interesting because most relevant theories primarily see incentives for lower instance courts to use the preliminary reference procedure, whereas they would expect last instance judges to be more reluctant in doing so. ⁸ The conducted interviews illuminate the different factors driving concerned national judges in deciding whether to refer preliminary questions to the ECJ. At the same time, they allow reflection about the role and significance of the obligation to refer which is to be found in the third indention of Article 267 TFEU and further developed in the well-known Cilfit-criteria. ⁹ More specifically, last instance court judges, in contrast to their lower instance courts colleagues, have an explicit obligation to refer preliminary questions when they have doubts about how EU law should be interpreted. This duty is implied by the plain language of the third paragraph of Article 267 TFEU, and it is backed by ECJ case law, with the Cilfit judgment which established the acte éclairé and acte clair exceptions to this obligation, ¹⁰ and which has been further enhanced and crystallized by means of subsequent cases. ¹¹ Thus, the gained empirical data allows to assess the relevance and power of the obligation to refer in Belgian highest court decisions of whether to refer.

Parallel to providing an insight into the different factors behind judges’ and court référendaires decisions about whether to refer, the conducted interviews also illuminate the possible reasons behind the exceptional attitude of the Belgian Constitutional Court towards the preliminary ruling procedure. Namely, the gained data enables an investigation into what would lead the country’s Constitutional Court to rank among the mechanism’s most frequent users in the EU. To be precise, the Constitutional Court has sent more than 140 different preliminary questions to the ECJ through more than forty preliminary references as of this writing, ¹² which makes the concerned Court an exceptional example of a very pro-active participant in the judicial dialogue with the ECJ. It has been suggested that the exceptionally proactive stance of the Belgian Constitutional Court could be explained by the fact that it was EC, later EU, law which in fact allowed the Court to gain the position of a full-fledged Constitutional Court. ¹³ This explains, according to Vandamme, why the dialogue between the Belgian Constitutional Court and the ECJ is so different than that of other Member States’ Constitutional Courts and why it is not startling that the Belgian Court displays an open attitude towards EU law. ¹⁴ Dyevre, moreover, suggests that the Belgian Constitutional Court uses the judicial dialogue with the ECJ as a means to evacuate sensitive cases concerning Belgian federalism. ¹⁵ However, Alen and Verrijdt, former and current judges at the Belgian Constitutional Court, disagree with this suggestion. ¹⁶ The authors hold that the unique attitude of the Belgian Constitutional Court can be explained by a set of factors including the idea that an openness towards EU law allows the Belgian Constitution’s obsolete fundamental rights catalogue to be updated. ¹⁷ The foregoing happens by means of strategic litigation by parties who, in light of the Belgian Constitutional Court’s openness towards the judicial dialogue, choose to litigate before that very court to get their preliminary questions submitted to the ECJ. ¹⁸ The interviews with Belgian judges provide for several interesting insights that allow to reflect on the above-mentioned suggestions made in the literature.

The discussion moves along as follows. First, the general overview of academic discussion regarding the various theoretical streams explaining judicial behavior in the context of the preliminary ruling procedure are briefly addressed in Section B. Second, the methodological approach applied in this study is explained in Section C. Third, the main results of the empirical study are presented in Section D. Finally, the match of the empirical results with the above “grand theories” is explored substantially in Section E. The Article ends with a general conclusion.

B. Grand Theories on the National Judges’ Engagement with the Preliminary Ruling Procedure

For a long time now, the motivations for national judges’ engagement in the judicial dialogue with the Court have been the focus of an intriguing and elaborate academic debate. ¹⁹ Indeed, there are many different “grand theories” that attempt to explain the phenomenon of national judges’ participation in the process of legal integration in Europe. While some scholars focused on the issue of differences in cross-national variations in preliminary references, ²⁰ a portion of the pertinent scholarship focuses on the potential causes of judges’ reluctance to use the mechanism of preliminary ruling. ²¹ Because different scholars provide excellent overviews of the existent theories in the field, only cursory attention will be paid below to the most relevant of them.

Legalism, neo-realism, neo-functionalism, the inter-court competition theory, and the individual profiles theory are among the major theoretical explanations for national courts’ involvement in the judicial dialogue with the ECJ. ²² The legalists suggest that the reasons behind preliminary questions are related to legal considerations—for example, the authority of the ECJ, the quality of its reasoning, and the simple fact that judges need to know how the law should be interpreted in order to apply it cause judges to resort to the ECJ. ²³ Called “most formal and hence the most naïve,” ²⁴ the explanation theorizes that a refusal to apply ECJ jurisprudence can be traced back to “unintended mistakes.” ²⁵ That is to say, only misunderstandings or a lack of knowledge would lead national judges to disregard the obligations following from EU law. ²⁶ Nuanced legalism, which constitutes a stream within the legalist theory, emphasizes the essential significance of legal arguments while simultaneously acknowledging that such legal arguments must be placed in the broader social and political context. ²⁷ Neo-realists, in turn, suggest that national political concerns and interests influence the involvement of national courts with the preliminary ruling procedure. Put differently, a calculation of the national political and economic interests affects the way judges make use of the preliminary ruling procedure. ²⁸ Neo-functionalists focus on the self-interests and motivations of individual judges and other legal actors in explaining their participation in European integration. The theory implies that the ECJ gave various incentives to various participants in the process, encouraging them to engage in judicial dialogue with the Court. According to Krommendijk, “the procedure is used as a ‘sword’ against the other branches of legislature or the executive.” ²⁹ Therefore, national judges reach for the preliminary reference procedure for a variety of individual reasons that, in general, relate to “judicial empowerment.” ³⁰ The judicial empowerment explanation focuses on “the extent to which judges work to enhance their own authority to control legal (and, therefore, policy) outcomes, and to reduce the control of other institutional actors, such as national executives, parliament, and other judges, on those same outcomes”. ³¹ The fourth theoretical stream of inter-court competition, which mostly derives from Alter’s works, assumes that national judges use preliminary rulings to advance their interests in the process of “bureaucratic struggles” between various levels of the judiciary. ³² Thus, it is important to consider the preliminary references in the context of competitive strategy involving various courts, or court levels. Finally, a more recent empirical research stream places emphasis on the role and influence of judges’ personal profiles and characteristics as well as the institutional setting in which they work with regard to how they participate in the process. By using empirical data from studies among judges from various EU member states, scholars contend that a variety of factors, also sometimes referred to as structural or technical factors, ³³ related to, for example, judges’ knowledge of EU law, experiences with applying it, or attitudes towards the EU, may affect how they use—or do not use—the preliminary ruling mechanism. ³⁴ In a similar vein, it is also suggested that the operational context in which judges function can exert a profound impact on the way they apply EU law and resort to the mechanism. ³⁵ In this regard, a number of variables, such as court resources, court specialization, or court culture, workload, and, closely related, time constraints, are listed as potentially influencing the willingness of judges to employ the procedure and the likelihood that questions will be referred to the Court. However, it should be noted that the many authors frequently disagree with one another. For instance, more recent evidence by Dyevre, Glavina & Atanasova have not supported the claim that jurisdictional specialization may lead to more preliminary references. ³⁶ All in all, academics seem to agree that a plethora of various legal and extra-legal factors may motivate judges to (not) participate in the preliminary ruling procedure, ³⁷ and that a general, coherent theory explaining the factors influencing judicial behavior in the concerned process is still lacking. ³⁸

When looking more systematically at the different factors that are put forward in the scholarship that aim to explain judges’ engagement with the mechanism, Krommendijk’s categorization is a very valuable one. ³⁹ The author separates the factors that may affect the way judges approach the procedure into three levels: Macro (the overall level of national judiciary), meso (court level), and micro (individual judge level). The figure underneath presents a general overview of available scholarship in which the different levels variables are used to explain possible reasons behind judges’ behavior in the framework of the preliminary ruling procedure (Figure 1).

Figure 1. Systematization of different motivations behind preliminary references.

C. Methodological Approach

This Article employs novel qualitative data gained in semi-structured interviews with Belgian highest national court judges and court référendaires to address the question regarding the reasons behind decisions about whether to make preliminary references to the ECJ. The qualitative interview data was collected by Kappé for the purpose of a Legal Research Master thesis project executed at Utrecht University and successfully completed in January 2023. Semi-structured interviews, in Galletta’s words, “create openings for a narrative to unfold, while also including questions informed by theory.” ⁴⁰ While being informed by theory, semi-structured interviews therefore allow for the further development of initial theories or, potentially, for the formation of a new conceptual framework or theory. This was especially desirable in the context of this research, as most of the above-mentioned theoretical streams explain primarily why lower-instance courts do not refer questions to the ECJ. It was therefore important to create space for a narrative on why last instance courts participate in judicial dialogue with the Court.

The use of interviews does have some drawbacks, though. First, the respondents may present a more positive image of themselves and/or conceal the real reasons behind their choice to refer or not. ⁴¹ Moreover, the selection of the respondents relies on the willingness of the judges and court référendaires to participate in interviews. This may result in bias in the research, as Leijon and Glavina have also noted, because involvement is more likely when the respondents have particular qualities, such as in-depth familiarity with EU law or a favorable attitude toward EU law. ⁴² The results of the interviews must therefore be interpreted with some caution.

Interviewees were selected using a non-probability sampling technique—more specifically, heterogeneous purposive sampling. ⁴³ In heterogeneous sampling, the researcher looks at a sample from multiple available angles, with the aim of achieving a greater understanding of the full “population.” ⁴⁴ Interviewees were therefore chosen in light of their possessed qualities, namely being a judge or a court référendaire at one of the Belgian highest national courts, which are the Constitutional Court, the Court of Cassation and the Council of State. The reason to include court référendaires in the sample was motivated by the fact they are very much involved in the judicial processes. They, for instance, analyze and summarize the procedural documents of the parties, and they may write the draft judgments. In that regard, court référendaires are closely involved in the decision whether to refer preliminary references to the ECJ and could provide immensely invaluable insights in that respect.

In total, ten interviews were conducted, of which six were with last instance court judges, which will be referred to in the footnotes as “J”, three were with court référendaires, which will be referred to in the footnotes as “R”, and one was with an auditor at the Council of State, which will be referred to in the footnotes as “A.” Like court référendaires, auditors do research on the cases that come before the Council of State, but instead of writing a draft judgment their role is more independent; they write advice to the Council of State which can then either be followed or not. Since the population of judges, court référendaires, and magistrates working for Belgian last instance courts is relatively small, ⁴⁵ ten interviews constitute a sufficient number to generate valid empirical findings on last instance court judges’ and law clerks’ motivations to refer.

Following the conduct and transcription of the interviews, “open” and “axial” coding were used to analyze the raw transcriptions. ⁴⁶ In order to identify the fragments, the interview transcription was first read and reread. If those fragments were considered relevant to the research, a code was assigned to it. Having coded all fragments using open coding, axial coding was employed. Axial coding involved firstly the determination of whether the codes developed in the process of open coding cover the data sufficiently and the creation of new codes if that was not the case. It was also checked whether the fragments had been coded properly, and if not, a new code was assigned. Then, where necessary, the fragments assigned to a specific code were subdivided. Lastly, the codes were placed in a coding tree, giving an overview of how the codes relate to each other. This coding tree served as the foundation for developing the next section that generally presents the findings.

D. Coercive and Persuasive Forces Behind Preliminary Questions

I. Coercive Impact of Article 267 TFEU and the Cilfit-Criteria: The Obligation to Refer

Interviews with the concerned judges and court référendaires provided insight into the factors that may affect decisions to refer or not to refer preliminary questions to the ECJ, as well as the importance of the Cilfit-criteria and the legal obligation to refer as outlined in the third indent of Article 267 TFEU. In fact, a resounding majority of respondents stated that they refer preliminary questions because it is an obligation placed on them by EU law. ⁴⁷ In that sense, the judges desire clarification on EU case law from the ECJ that would subsequently help them resolve or avoid legal uncertainty. The following quote exemplifies it well:

If when reading a text, we find there is an interpretation question… and if no case law is found on it that indicates the correct solution, we can ask a question to the ECJ also to give the ECJ the chance to give that clarification. … It is not for us to be the very first to give a clarification of a rule that is not so clear. ⁴⁸

Two respondents indicated that the decision on whether to refer was based exclusively on legal considerations which stem from Article 267 TFEU and the Cilfit-criteria. ⁴⁹ It was held that: “[T]o propose [to make a preliminary reference] and for the judges… to ask, it is a solely legal decision based on Article 267 of the Treaty of the Functioning of the European Union and the Cilfit case law.” ⁵⁰

When asked about the underlying reason for the adherence to the rules as enshrined in Article 267 TFEU and ECJ’s jurisprudence, respondents indicated that they wished to be loyal to the system of EU law in light of its functioning and democratic legitimacy and because “it is one of the basic principles of the rule of law that one applies the law.” ⁵¹

Inter-court dynamics between the highest Belgian courts and the ECJ, as well as those between the highest Belgian courts and the lower Belgian judiciary, were also cited as motivation for adhering to the legal obligation to refer. In relation to lower courts, some felt that the right example needed to be set in not placing themselves as the highest national court above the rules of law. ⁵² In relation to the ECJ, one respondent indicated that abiding to the legal obligation to refer allowed the considered court to “speak the ECJ’s language” and “create goodwill,” enabling it to maintain a genuine, critical, dialogue with the ECJ. ⁵³

Beyond the applicability of Article 267 TFEU and the Cilfit-criteria, several respondents, however, also indicated additional legal or extra-legal grounds influencing the decision whether to refer. Importantly, it was emphasized that these grounds were considered in interpreting the legal obligation to refer set forth in Article 267 TFEU and were practically never cited as “autonomous” reasons for making or not making a preliminary referral. The remainder of this section will firstly point out legal and extra-legal reasons of Belgian highest court judges and référendaires to refer. Subsequently, legal and extra-legal reasons not to refer will be set out. Given the fact that legal and extra-legal reasons frequently intertwine, they will be addressed together in the same section.

II. Persuasive Legal and Extra-Legal Reasons to Refer

Besides receiving clarification regarding EU law from the ECJ through the preliminary ruling procedure, judges also indicated a challenging function of the judicial dialogue. In that latter case, judges for instance make a preliminary reference to probe a certain line of the ECJ’s case law that is perceived as rather strict by the national court. An example in which existent case law of the ECJ formed a reason to refer, is a referral on data-retention by the Constitutional Court from 2018. ⁵⁴ It was explained that this referral was made because in earlier case law the ECJ interpreted the data retention directive ⁵⁵ too strictly in the judge’s opinion. It was held that “[i]f the intelligence services themselves do not have the permission to make use of the possibilities offered by technology, then they are not playing on an equal footing and then it is impossible to avoid the next terror attack or to fight the mafia.” ⁵⁶

The judge therefore made a preliminary reference to challenge the rather strict line of case law of the ECJ “because of crime-fighting and counter-terrorism interests.” ⁵⁷ The above quote also shows how legal and extra-legal considerations to make a preliminary reference may intertwine. The legal reason of questioning a certain line of ECJ’s case law interacts with an extra-legal reason of making a reference, which is underpinned by national public interest in a hope to obtain a less strict interpretation by the ECJ. This, in turn, would allow the Belgian government to better address the problem of crime and terror. In a similar vein, a reason mentioned by several respondents was the wish to have a certain interpretation of the law apply for the whole of the EU. ⁵⁸ It was indicated that this was either for reasons of uniform application of EU law—which we could classify as a legal reason—or to have a certain interpretation of EU law apply for the whole of the EU. This, in result, would protect—or not disadvantage—the Belgian State or Belgian residents compared to other Member states and their residents. This is persuasively reflected in the following citation:

We protect the Belgian government [by making a preliminary reference to the ECJ], because if we ourselves … were to give a very far-reaching interpretation of European law that was very detrimental, financially or administratively, to the Belgian state, only Belgium would be bound by that, because we do not bind the other Member States. Whereas if we let the Court of Justice give that interpretation, then every Member State suffers equally. Then we do not disadvantage Belgium in relation to the other Member States. ⁵⁹

Concerning protection of Belgian residents, a Constitutional Court judge held that:

Conversely, of course, we also protect Belgian residents. It could just as well be that all the other Member States have transposed a directive correctly and Belgium has not. Then Belgian residents or companies wishing to establish themselves in Belgium are at a disadvantage compared to nationals and companies from all the other Member States. ⁶⁰

Another intriguing reason than can persuade judges to refer that was put forward could broadly be referred to as a fear of “costs and loss of face.” This phenomenon would result from either a conviction of Belgium by the ECJ in an infringement procedure for incorrect transposition or violation of EU law, a conviction of the highest court for a refusal to comply with the obligation of last instance courts to request a preliminary reference, a conviction by the European Court of Human Rights for violation of the right to a fair trial or, finally, from the establishment of state-liability which could be incurred due to an infringement of last instance courts’ duty to refer. ⁶¹ Furthermore, in connection to the establishment of state-liability, one respondent stressed the role of the parties in incentivizing judges to refer a question. The judge held, “if there is a party that threatens with liability proceedings, one might be quicker to say: yes, there is a sufficient doubt that justifies referring questions to the Court in Luxemburg.” ⁶²

Closely related to the foregoing, respondents also indicated that when parties strongly insist on a preliminary reference, this may constitute a reason to refer. ⁶³ In addition, the impact of the case, which may relate to either societal impact, economic impact on the parties, or impact on the development of EU law was pointed out as a likely reason to resort to the preliminary ruling procedure. ⁶⁴ This is adequately illustrated in the following quote: “Indeed, I think the specificity of the legal question or the broader fallout that it may have, is something that comes into the consideration as well [in the decision not to refer]. Not explicitly, of course, but that’s something that you take into the back of your mind.” ⁶⁵

Respondents furthermore indicated that in some situations, they would rather have the ECJ decide on a certain issue. ⁶⁶ This can be motivated firstly for reasons that could be classified as blame avoidance or a strategy to avoid responsibility. One respondent held that when a norm stands a good chance of being declared unconstitutional,

[T]he Court wants to cover itself when Union law is at stake, so to speak, by putting the balls to the Court of Justice and then saying ‘yes but we had no choice but to declare that regulation unconstitutional or annul that regulation’ …sort of to not have to take the responsibility entirely on its own shoulders in relation to the government. Yes, I do think that happens. ⁶⁷

It is yet important to emphasize that when other respondents were asked whether strategic motivations could play a role in their decision whether to refer, no support was given to the “blame avoidance strategy,” nor did they indicate that any other strategic considerations have a bearing on the decision whether to refer. A different motivation for judges to rather have the ECJ decide on a specific case mentioned in the interviews was the fact that an answer from the ECJ could settle an internal disagreement within the national court on how the case should be resolved. ⁶⁸ In that sense, judges might consider shifting the responsibility for resolving a difficult case to the ECJ so that they do not have to resolve internal disagreement between judges on how to decide the case. Contrastingly though, another judge expressed a disagreement with the suggestion found in the literature that the Constitutional Court sometimes asks a preliminary question in order to not have to resolve sensitive cases themselves. A final reason to have the ECJ decide on a certain issue indicated by respondents was the view that, on certain delicate topics, it can be useful to make a preliminary reference to give more weight to the judgment. ⁶⁹

The final extra-legal drive behind referrals that needs to be noted here has to do with judges’ openness to EU law and overall willingness to refer. Some respondents confessed to having “a European reflex,” ⁷⁰ or to try “to do well in the European class.” ⁷¹ When questioned about the fundamental cause of their positive receptivity to EU law, respondents mentioned relevant structural factors as playing a role. More precisely, knowledge of EU law, either of the individual judge or among the support staff, ⁷² and other operational factors such as availability of databases including analysis of EU (case) law. ⁷³ Also, the frequency of exposure to queries regarding the interpretation and validity of EU law, being a structural factor, constitutes a possible reason worth highlighting. ⁷⁴ The EU law exposure can be enhanced by the substantive field of law the judges worked in when it is characterized by a high degree of EU harmonization such as VAT or customs, public procurement or environmental law. Importantly, the role of the representatives of the parties in identifying an EU law-related question was viewed as significant in these highly Europeanized areas of law. According to one of the respondents: “I would say that the role of the lawyers is very important in the matter of public procurement. I would almost say that if the lawyers don’t identify the problem, then there probably won’t be anything to identify.” ⁷⁵ This important role played by legal representatives of the parties was explained by reference to the quality of their legal expertise in the concerned sectors of law. It was held that due to the very technical nature of public procurement, and similarly tax law, the quality of the lawyer’s expertise is high, and they follow the new developments in EU law very closely. ⁷⁶

Regarding the unique openness of the Constitutional Court towards the preliminary ruling procedure, one respondent stated that this attitude was rooted in the old strategic decision to use EU law to become a full-blown Constitutional Court:

I think the strategic behaviour of our Court is somewhere old in the sense that we have our jurisprudence which is based on Articles 10 and 11 of the Constitution, through which . . . our Constitutional Court has a very low threshold open for direct review of EU law and subsequently also to referring questions to the Court of Justice, much more open than most other constitutional courts. So I think you could possibly say that that was a sort of yes strategic decision, which is very old indeed.… So hence we actually have a huge openness to international and European law, which of course has greatly increased the Court’s scope of review. And there has certainly been some tension there with the Supreme Court. ⁷⁷

Another court-specific legal reason to refer mentioned by Constitutional Court respondents is that the judicial dialogue with the ECJ allows the Belgian Constitution’s obsolete fundamental rights catalogue to be modernized. ⁷⁸ This was considered both a motive for the general openness to EU law, as well as for their willingness to engage in the judicial dialogue with the ECJ. ⁷⁹

Finally, it is worth stressing that a Constitutional Court judge indicated that the respective court was more often confronted with cases including a question about the interpretation or validity of EU law due to the flexible procedural rules regarding locus standi and the open attitude of the concerned judges towards the dialogue with the ECJ. This favorable procedural framework regarding parties’ access to the Constitutional Court that is found in Belgium combined with the open attitude of respective judges, results in an absorbing phenomenon reflected in the following quote: “Applicants from all over Europe have now discovered [the Belgian Constitutional Court] as a mechanism to bring their problem of European Union law before the Court of Justice.” ⁸⁰

III. Persuasive Legal and Extra-Legal Reasons Not to Refer

The respondents also pointed to possible reasons that may keep judges away from the preliminary ruling mechanism. Besides the exceptions to the obligation to refer created in Cilfit, the length of the preliminary reference procedure (in extremely urgent procedures) was stressed. ⁸¹ As explained by a respondent from the Council of State, in extremely urgent interlocutory procedures, a judgment must be pronounced within 45 days after the administrative decision was contested. In light of the duration of a preliminary reference, which on average takes approximately 15 months, ⁸² referring a question to the ECJ would therefore not constitute a workable solution. Closely related to that, the time investment necessary to ask a preliminary question was emphasized as a reason not to refer. The need for speedy administration of justice for the parties, as well as workload concerns shared by judges, are the root causes of this. To decide a dispute, a judge would have to re-examine the relevant case once the ECJ issues its ruling. One of the respondents went as far as to claim that the duration of the preliminary reference procedure constitutes the most important consideration in a judges’ decision not to refer:

[A]n important element that comes into play is the length of the process. Knowing that if you put the question to Luxembourg, you are easily a year, two years further before you can finally make a judgment and that the process takes that much longer. That’s surely a consideration to stick to that line, of not asking the question at the slightest doubt and maintaining a somewhat broader and flexible [interpretation of the Cilfit-criteria]. That’s certainly an important consideration. I think maybe the most important one, actually. ⁸³

Interestingly, next to the additional time investment, additional costs—such as extra lawyers’ fees—induced by the parties if a question is referred were also indicated to be considered in the decision whether a reference should not be made. ⁸⁴ Respondents also mentioned the workload of the Court of Justice as a reason that can make them somewhat reluctant to make preliminary references, especially when it concerns cases that are “not very interesting” for the ECJ. ⁸⁵ In that regard, it was pointed out that where the interest at stake is very minor, or where the question of law is highly specific, judges and law clerks may be less inclined to make a reference.

Another (legal) reason not to refer is related to the passive role assigned to judges of the Court of Cassation in civil cases which does not allow judges to go outside the scope of the ground of appeal and highlights the key role of parties to the proceedings in bringing preliminary questions to the ECJ. In those situations, a preliminary reference will almost exclusively be made when requested by one of the parties to the dispute. Consequently, if no preliminary reference is requested by the parties, referring it of the judges’ own motion might easily go outside the scope of the ground of appeal.

The quality of the judicial dialogue was mentioned by one of the respondents as a possible barrier to refer. ⁸⁶ The respondent brought up the possibility that the ECJ’s relatively limited engagement and interaction with the reasons put forth by the last instance court in a preliminary reference might constitute a reason not to refer a question. It was held that:

[A]nother obstacle is of course if in the judicial dialogue there is little response or that the arguments we raise are not really interacted with.… So if, in that area, there is little interaction with the arguments that are raised, that might be a reason to be a bit more cautious in the interpretation of the [Cilfit] criteria. So not to disregard the criteria, but perhaps to be a little more cautious in asking the question. ⁸⁷

Possible disagreement with the evolution of the ECJ’s caselaw and, thus, with the anticipated response from the ECJ regarding a specific question, was another potential reason not to refer. One of the respondents cited a situation in which the national court did not use the possibility to “question the ECJ’s case law” because they disagreed with how the ECJ’s case law had developed in the relevant areas and, consequently, with the expected response to their question about the interpretation or validity of EU law and the possible consequences of it. ⁸⁸ In a similar vein, a judge emphasized hesitation to refer by observing that: “[Y]ou have a particular problem and you can then ask the ECJ a question, and you know what you put in, but God knows what comes out and what the consequences are. And we think about that a lot anyway.” ⁸⁹

A further reason not to refer mentioned in two interviews is that, when a reference is not strictly necessary, some judges would rather solve the case themselves. ⁹⁰ It was held, for instance, that: “If you have a problem that you would possibly ask a question about, but you don’t have to, the judge’s core business is solving cases. So if you can solve them, you solve them immediately.” ⁹¹

Similarly, another judge indicated that:

[W]hen the solution seems sufficiently clear,… you might then also prefer to solve [the case] yourself and not hand it over to a third body—Court of Justice in this case—who can then say in your own place what the solution actually is. That doesn’t play as strongly, but you can assume that if the solution seems clear, then that will be able to be an obstacle. If the solution does not seem clear, of course that does not play a role. ⁹²

Lastly, a few judges noted that the choice not to refer was influenced by reputational issues related to asking a question that was deemed “silly.” The narratives displayed that judges might feel like they asked a “silly question” when the ECJ responds to their preliminary question with an order rather than a judgment, implying that the question posed was actually an acte claire or an acte éclairé, or when the ECJ asks the referring court additional questions before handing down its decision, such as whether the national last instance court is aware of a particular judgment by the ECJ and whether it would like to maintain their question in light of that judgment. In that respect, one judge held that: “I see that that is with me and with the people with I have always worked, and I suppose with every judge: we don’t want to lose face with the Court of Justice because of a silly question either.” ⁹³

E. Unfolding Story of Judicial Dialogue in the EU but No New Grand Theory

This section will go through how the empirical data gained from interviews fits with the theoretical framework presented in Section B. It is worth emphasizing that during the interviews, participants were specifically questioned about their own understanding and agreement with the respective “grand theories.” The evaluation of these responses aids in determining whether and to what extent the concerned judges might be driven by motives that are discussed in the literature. Subsequently, it will be briefly considered whether the possible reasons behind the exceptional attitude of the Belgian Constitutional Court mentioned in the introduction are supported by the present empirical findings. It is necessary to stress that the Court’s exceptionalisms did not stand central in the interviews as such. However, the gained data does allow reflection on the issue to some extent.

It does not come as a surprise that the empirical findings from the interviews are to a high extent aligned with the stream of—nuanced—legalist theory which gains our strongest support here and further confirms findings by Krommendijk, ⁹⁴ Leijon and Glavina, ⁹⁵ and Jaremba. ⁹⁶ According to few respondents, those were legal(ist) factors alone that determined whether or not to refer, supporting the mainstream legalist theory. The majority, however, indicated that next to the legal obligations to refer stemming from Article 263 TFEU, a number of extra-legal motivations and structural factors may play a role in the process. The empirical results provide strong evidence that judges and court référendaires consider for instance the length of the preliminary reference procedure and the necessary time investment when deciding whether to refer a case. The length of the procedure which could not only add to the workload judges struggle with but also slow down the administration of justice at the expense of the parties support previous findings by inter alia Jaremba, Pavone, Krommendijk, Glavina, and Leijon. ⁹⁷ Moreover, as also put forward by Krommendijk, ⁹⁸ and by Leijon and Glavina, ⁹⁹ concerns about the capacity of the ECJ to process preliminary references proved to be of relevance. This until recently unrecognized ground relates to “an attempt to abide by one of the most important professional responsibilities of a judge to safeguard the proper functioning of the legal system.” ¹⁰⁰ Next, the possible impact of the judgment provided by the ECJ to the national judge’s request is also found to be taken into consideration while deciding whether to refer. This finding again concurs with Krommendijk, ¹⁰¹ who argued that, particularly at the level of lower instance courts, “a reference is more likely where the question plays a role in a considerable number of cases or where the financial or societal consequences are substantive.” ¹⁰² This supports a thesis that such considerations may also play role at the higher level of judiciary.

The empirical findings provide somewhat modest support for the neo-realist theory. Several respondents mentioned that they would make a preliminary reference to have a specific interpretation apply to the entire EU. According to one reply, this was done to protect the Belgian state and/or its residents. Moreover, neo-realist motivations may occasionally be used as a strategy to contest the ECJ’s existent line of jurisprudence. None of the respondents, however, expressed concerns that national governments or politicians may use specific measures to restrict the range of appreciation for participation in European integration. Similarly, there was little evidence that judges and legal secretaries were hesitant to stray too far from their countries’ economic and political priorities out of concern for their governments’ potential negative response.

The neo-functionalist theory advanced mostly in the writings of Mattli and Burley (Slaughter) and which focuses on individual incentives and self-interest of different players—judges themselves, practitioners, individual litigants—to participate in European integration gains some support as well. ¹⁰³ The narratives indeed prove that these grounds may propel preliminary referral ahead at times. In particular, the empirical findings underscore the role and involvement of private parties in pushing for preliminary references. One respondent emphasized that preliminary references are frequently requested by parties, who sometimes even threaten courts with state liability procedures if the last instance court does not resort to the procedure. In another narrative, it was mentioned that the judge’s choice of whether to refer a question to the ECJ was affected by the degree to which parties insisted on a preliminary reference. Lastly, one respondent observed that in cases involving complex and heavily EU-law-influenced sectors, the lawyers even played an indispensable role in raising issues of EU law and the necessity to refer a preliminary question.

Furthermore, the empirical findings provide no support for the inter-court competition account. However, this conclusion should definitely not be generalized and at the same time it is not startling either. In the end, the empirical research was conducted among last instance courts which find themselves at the apex of national judiciary, the fact which introduces certain bias in the narratives. What is more, the inter-court competition theory is based on the assumption that the preliminary ruling procedure empowers lower national courts judges, a cause which is logically suppressed in case of higher instance court. At the same time, the inter-court competition model puts forward that national higher courts are likely less willing to employ the mechanism because they do not want to share their powers with the ECJ, ¹⁰⁴ a phenomenon which our data cannot confirm or even stands in contrast to.

Finally, the present project found some evidence for the exceptional attitude of the Belgian Constitutional Court towards the preliminary reference procedure. First, in line with suggestions done by Vandamme, ¹⁰⁵ it was noted that the openness of the Constitutional Court was rooted in a strategic decision to use EU law to become a full-blown Constitutional Court. Second, it was indicated that a reason for Constitutional Court judges to refer preliminary questions was that the judicial dialogue with the ECJ allowed to modernize the Belgian fundamental rights catalogue. These findings support Alen and Verrijdt who relied on a similar argument to explain the unique openness of the Belgian Constitutional Court towards EU law and the preliminary reference mechanism. ¹⁰⁶ Third, the unique and pro-European approach presented by the Constitutional Court has not go unnoticed, and is may now be used by parties who turn to that very Court for purpose of strategic litigation.

Having said that, several general conclusions can be drawn. First, even though the nuanced legalist theory gains most support in our case here and should be seen as a starting point for further discourse, almost all other theoretical streams provide for some interesting and valid insights and explanations that may be relevant at times. This leads us to a firm confirmation of earlier findings that no general, coherent “grand theory” valid for all national judges in the EU can be established.

Second, legal and extra-legal grounds to refer very much intertwine, intersect and at times it is nearly impossible to separate them from each other. Except for very explicit but exceptional narratives, it partly remains a matter of subjective interpretation to classify a motivation behind a (non)referral as a legal or extra-legal one. However, we conclude that for the Belgian last instance courts judges and court référendaires the legal factors—for instance, the legal and coercive force of Article 267(3) TFEU and relevant case law creating exceptions to the obligation to refer, are always the starting point for their action. If the judges make a decision not to refer based on exceptions established in Cilfit, then frequently non-legal motivations, which predominantly pertain to structural and institutional factors, will play a role in such a decision. According to Leijon and Glavina, extra-legal reasons not to refer are of particular relevance, because the existence of extra-legal reasons is problematic in light of the functioning of the preliminary reference procedure. Especially in the context of last-instance courts, the authors argue the absence of dialogue in specific cases may lead the integration process to “run into conflicts and misunderstandings.” ¹⁰⁷

Third, the more recent empirical stream into individual profiles of judges provides very valuable insights into the grounds and motivations behind preliminary references. In fact, practically all factors classified as meso and micro were mentioned in interviews at some point, albeit not in a systematic way. Finally, the factors which are classified by Krommendijk as macro were illuminated in the narratives in a less explicit and extensive way but also here we see some variables, for instance the adversarial/inquisitorial nature of the legal system, gaining confirmation.

F. Conclusion

The participation of national judges in the process of judicial dialogue with the ECJ and the possible causes and motivations behind such participation have been in the center of academic attention throughout the decades. Undoubtedly, the functioning of the preliminary reference mechanism is dependent on the willingness of national courts to take part in the process. Even though there have been many both theoretical and empirical attempts at explaining this participation or the lack thereof, the available data is still fragmented and covers only a limited number of national jurisdictions in the European Union. Moreover, the vast majority of the literature on judges’ use of the preliminary reference mechanism focuses on lower instance courts. It is against this backdrop that an examination of reasons (not) to refer of highest national courts judges is particularly interesting because most relevant “grand theories” primarily see incentives for lower instance courts to use the preliminary reference procedure, whereas they would expect last instance judges to be rather reluctant to do so. It was the very aim of this Article to partly fill in the void in the literature and to provide novel insight into the (non) participation of Belgian highest courts judges in the mechanism and the motivations and causes that may influence their decision to (not) refer. The data gained mostly speaks in favor of the nuanced legalist theory, but it must be stressed that almost all other theoretical streams provide for some relevant and valid explanations of judicial behavior vis-à-vis the preliminary ruling procedure. Concurrently, the present research facilitated to some extent our understanding of the possible reasons behind the very exceptional, pro-active attitude of the Belgian Constitutional Court towards the preliminary ruling procedure. Clearly, this unique approach of the Constitutional Court, and the reasons behind it, is unparalleled in other national jurisdictions and should be interpreted in the context of national legal developments.

This research fills the void in the literature only partly though. As our findings definitely cannot be generalized and used to explain judges’ behavior in other EU Member States, ¹⁰⁸ a suggestion for further research would be to conduct more research on reasons of whether last instance court judges refer in Member States, which is not yet covered in the literature. We still know very little about a number of EU jurisdictions, in particular in the newer member states of the Union, which might benefit from further academic investigation. Finally, this Article deepens our knowledge of the strengths and weaknesses of the preliminary reference mechanism and the dynamics attached to it, making it possible to conduct further research on how the preliminary reference might be improved in order to enhance the judicial dialogue in the European Union.