A. Constrained discretion to refer

National courts have the ‘widest discretion’ in referring matters to Luxembourg.Footnote ⁹ This freedom is an ‘inherent part of the system of cooperation’ established by the preliminary reference procedure.Footnote ¹⁰ Importantly, that discretion can be exercised at whatever stage of the proceedings judges consider appropriate.Footnote ¹¹ A slowly growing body of case law started to restrict this discretion in German courts. This applies especially to procedures governed by the Code of Civil Procedure (‘ZPO’). Matters below a certain monetary threshold are adjudicated in first instance at the ‘Amtsgericht’, whereas those crossing the threshold come before the regional court (‘Landgericht’). The latter is divided into different chambers of three judges. Yet, for reasons of efficiency cases are usually decided by a single judge. If the case is of ‘fundamental significance’, the respective judge must exceptionally submit the dispute to his or her chamber.Footnote ¹² According to a recent line of jurisprudence, this might be the case when a judge intends to submit a preliminary reference.Footnote ¹³ If the single judge disregards this obligation, he might even face a motion for recusal due to a flagrant breach of procedure.Footnote ¹⁴

The compatibility of this mechanisms with Article 267 TFEU has reached the Court of Justice in several instances. Even though the Court eventually refrained from assessing this practice,Footnote ¹⁵ Advocate General Rantos stressed that Article 267 TFEU precludes such a rule. In this context, he reemphasised that the national court must remain ‘free to refer to the Court for a preliminary ruling any question that it considers necessary, at whatever stage of the proceedings it considers appropriate.’Footnote ¹⁶ Along these lines, further references raised this ‘disconcerting development’ to the attention of the Court but were later removed from the Court’s register.Footnote ¹⁷

In parallel, it has become highly controversial, whether there are remedies against the decision to refer. This might place an additional limitation on the judges’ discretion. Generally, German law allows to challenge decisions and orders of courts, which are no judgments, through complaints.Footnote ¹⁸ Nevertheless, it remains controversial whether this applies also to preliminary references. The case law of the Court is ambiguous in this respect and is quoted both by those advocating and opposing the possibility to appeal such decisions.Footnote ¹⁹

To start with, it is established jurisprudence that decisions referring a question to the Court can remain subject to the remedies normally available under national law.Footnote ²⁰ In Cartesio, however, the Court stated that national law cannot permit higher instance courts to set aside decisions to refer or to order the resumption of national proceedings while awaiting the return of the preliminary reference.Footnote ²¹ At the same time, the Court did not consider that appeals against decisions to refer were incompatible with Union law per se. In more recent case law, it stressed that

At first glance, there are indeed good reasons for admitting such remedies. Referrals have an immediate impact on the parties. They imply a considerable delay, additional submissions, and therefore higher costs. Accordingly, referring the case to Luxembourg might not always be in the parties’ interest. In this spirit, some argue that – even if not the referral itself – the decision suspending the national proceedings should be subject to remedies.Footnote ²³ The court seized to rule on such a complaint could assess, whether the reference is based on a manifestly erroneous assessment of the facts or the applicable law.

Nevertheless, such a distinction between the decision to refer and the decision to stay the proceedings will be difficult to maintain. A reference without suspending the main proceedings is hardly feasible and not in line with Article 23 of the Court’s statute. As such, lifting the decision on the suspension will have an immediate bearing on the referral.Footnote ²⁴ In case of administrative, financial, and criminal procedural law, many argue that decisions to refer should be considered as mere ‘procedural guidance’, which are specifically excluded from complaints in the respective procedural codes.Footnote ²⁵ But also in civil procedures, where such an exception does not exist, many higher regional courts have rejected the admissibility of complaints against the decision to refer.Footnote ²⁶ First, these courts refer to the functioning of the preliminary reference procedure, which is not only an individual remedy but also an inter-court dialogue seeking to guarantee the uniformity of EU law. This might compel judges in certain situations to refer also against the express will of the parties. Second, the cooperation under this mechanism requires that national courts are as free as possible to refer, at any stage of the proceedings, any question they consider to be necessary. Remedies would severely circumscribe this discretion.Footnote ²⁷ And third, the complaint against the decision to suspend proceedings seeks preventing an unjustified standstill. Yet, preliminary references prepare and foster the decision-making process in the main proceedings.

While there was some considerable disagreement between the higher regional courts,Footnote ²⁸ the Bundesgerichtshof has left this issue explicitly open.Footnote ²⁹ As such, it remains to be seen which position will carry the day.

B. Obfuscated duty to refer

According to the Court of Justice, the duty to refer under Article 267(3) TFEU does not only comprise those courts of last instance, against whose decisions there is generally no appeal (abstract view), but also to decisions, which are – in the specific case – not subject to further remedies (concrete view). In this sense, the Court embraced the idea of ‘concrete appealability.’Footnote ³⁰ With regard to the German legal system, this makes things quite tricky. Indeed, the procedural framework fails to allocate the obligation under Article 267(3) TFEU in any clear and predictable manner. Instead, it establishes a complex and confusing system, which fosters a diffusion of responsibility.

Starting with the abstract view, all federal courts, such as the Federal Court of Justice (Bundesgerichtshof), the Federal Administrative Court (Bundesverwaltungsgericht), the Federal Fiscal Court (Bundesfinanzhof), the Federal Labour Court (Bundesarbeitsgericht), and the Federal Social Court (Bundessozialgericht), as well as the constitutional courts at the state and federal level are subject to the duty in Article 267(3) TFEU.Footnote ³¹ In this context, it should be stressed that constitutional complaints before the Bundesverfassungsgericht do not constitute a ‘remedy’ in the sense of Article 267(3) TFEU.Footnote ³² Since such a complaint is also possible against judicial decisions (so-called ‘Urteilsverfassungsbeschwerde’), even the federal courts would be excluded from the duty to refer. Such an interpretation would hardly meet the aim of Article 267(3) TFEU and severely undermine the cooperation among Luxembourg and the German judiciary.Footnote ³³ As such, all federal courts come under the duty to refer.

Switching to the concrete view muddles this picture. Under German procedural law, judgments can generally be appealed on factual and legal grounds in second instance (so-called ‘Berufung’) and on legal grounds in third instance (so-called ‘Revision’). Especially in civil and administrative proceedings, however, the court, whose decision is to be appealed, must in many cases admit the appeal.Footnote ³⁴ In some cases, the decision refusing to admit an appeal can be challenged with a complaint. This leads to a highly complex picture when it comes to determination of who is subject to the obligation under Article 267(3) TFEU.Footnote ³⁵

To provide an example, first instance judgments by civil courts can only be appealed once they reach a certain monetary threshold (600 EUR). In all other cases, the appeal must be specifically admitted by the court of first instance.Footnote ³⁶ When the threshold is not met and appeal not granted, there is no further possibility to challenge the judgment. In consequence, the first instance court becomes a court of last instance under Article 267(3) TFEU. Similarly, second instance courts must admit the appeal in third instance.Footnote ³⁷ However, their decision to deny an appeal can be challenged before third instance courts.Footnote ³⁸ In this complaint procedure, the third instance court can review whether the conditions for granting an appeal, such as the ‘fundamental significance’ of the case, are met. This is usually the case when unresolved questions of EU law arise. For that reason, the complaint procedure is considered a ‘remedy’ under Article 267(3) TFEU.Footnote ³⁹ If the conditions for such a complaint are met (the monetary threshold is 20,000 EUR), the duty of Article 267(3) TFEU shifts to the third instance, if not, the second instance court is obliged to refer.

Confused? I wouldn’t be surprised! Even well-versed practitioners get caught up in the intricacies of this system. Indeed, it is difficult to understand, for both litigants and judges, when the obligation to refer actually bites. Many judges might thus disregard their obligation to refer in their daily practice. First instance judges in civil procedures, for example, may infringe Article 267(3) TFEU by neither referring nor allowing the appeal in cases that do not reach the value of 600 EUR. Such day-to-day cases can still raise contentious issues of EU law, such as questions of consumer protection or passenger rights. This is all the more relevant as the parties in such cases do not require legal representation. Similar infringements can arise in complaint procedures against decisions to not admit an appeal, which lead to a back-and-forth between second and third instance. Here, litigants might face a diffusion of responsibility for the obligation to refer between different levels of courts.Footnote ⁴⁰

This is a problem. Especially under Article 267(3) TFEU, the preliminary ruling mechanism is a further layer of effective judicial protection in the service of the individual.Footnote ⁴¹ Even if there may be good reasons for this complex appeal system, it seems ill-equipped for the smooth operation of the preliminary reference mechanism.

This challenge to the duty to refer under Article 267(3) TFEU is further aggravated by the lack of sufficient remedies to enforce this obligation. Leaving remedies under European law, such as state liability,Footnote ⁴² infringement proceedings,Footnote ⁴³ or applications before the ECtHRFootnote ⁴⁴ aside,Footnote ⁴⁵ three possible remedies, ordinary, extraordinary, and constitutional, might be envisaged. To start with, there are little to no possibilities for ordinary appeal. Most cases under Article 267(3) TFEU will concern jurisdictions of last instance. This leaves the few cases in which lower courts have a duty to refer, such as those under the Foto-Frost line of jurisprudence. These could be appealed in higher instances.

Second, some suggest employing an exceptional remedy granted in the event a party’s right to be heard has been disregarded.Footnote ⁴⁶ This remedy allows applicants to continue proceedings against which no legal remedies are available if the respective court has infringed the parties’ right to an effective and fair legal hearing guaranteed by Article 103(1) of the German Basic Law. This remedy is raised before the iudex a quo, meaning the judge or court whose decision is challenged. Initially, the Bundesgerichtshof contemplated whether this remedy could be applied also to other procedural rights, such as the right to a lawful judge under Article 101(1) of the Basic Law.Footnote ⁴⁷ Infringements of the duty to refer under Article 267(3) TFEU could have been addressed in this way. Being a highly exceptional remedy, however, this procedural tool cannot be extended beyond its scope of application. As such, it is limited to the right to be heard.Footnote ⁴⁸ Subsequently, this narrow interpretation has permeated the case law of German apex courts.Footnote ⁴⁹

Third, disregarding the duty to refer under Article 267(3) TFEU can give rise to a constitutional complaint based on the right to a ‘lawful judge’ under Article 101(1) of the Basic Law.Footnote ⁵⁰ That provision guarantees that individuals are not deprived of the court or judge with jurisdiction over the respective proceedings. This right does not only encompass national courts but also the Luxembourg judges. Thus, a court of final instance that disregards its duty to submit a preliminary reference infringes a constitutional right.Footnote ⁵¹ Nevertheless, the Bundesverfassungsgericht does not understand itself as ‘oberstes Vorlagegericht.’ Instead, it discharges this task with outmost restraint and only reviews whether the handling of Article 267(3) TFEU is ‘manifestly untenable.’Footnote ⁵² In this context, Karlsruhe identified three cases in which such a situation might arise.Footnote ⁵³ The first situation concerns a fundamental disregard of the duty to refer. Although the respective court recognises that EU law is relevant for the resolution of the case before it and demonstrates doubts as to its correct interpretation or application, it does not submit a reference. The second scenario concerns the deliberate deviation from the case law of the Court without referring. The third and most controversial case relates to situations in which the respective judges face an incomplete case law of the Court.Footnote ⁵⁴

Initially, Karlsruhe’s two senates handled the third category differently. More leniently, the second senate refused to fully review the national court’s obligation under Article 267(3) TFEU and simply assessed whether a different interpretation of the respective EU law from that chosen by the national court is ‘clearly preferable.’ Put differently, Article 101(1) of the Basic Law is not infringed, when the national court’s interpretation is ‘at least justifiable.’Footnote ⁵⁵ Much stricter, the first senate assessed whether the national court handled its duty to refer under Article 267(3) TFEU, including the CILFIT-criteria, in a reasonable manner.Footnote ⁵⁶ In particular, it must provide reasons as to whether and why the question is an acte claire or an acte eclairé. Under the pressure of the academic literature,Footnote ⁵⁷ the second senate followed suit and aligned its jurisprudence.Footnote ⁵⁸

Some scholars have argued that ‘this may be one of the most efficient enforcement mechanisms with respect to the duty to make preliminary references.’Footnote ⁵⁹ Yet, this impression can hardly withstand closer scrutiny. First, we can assume that only a very small portion of constitutional complaints is successful. Between 2010 and 2023, the Karlsruhe court has decided 118 cases concerning a violation of the duty to refer. Out of these, it ascertained a violation in 29 cases.Footnote ⁶⁰ While this success rate of 25 per cent seems like a relatively strong outcome, it must be noted that the Bundesverfassungsgericht usually rejects between 97 and 98 per cent of its cases through a so-called ‘Nichtannahmebeschluss’,Footnote ⁶¹ an order refusing to admit the complaint which is usually neither reasoned nor published.Footnote ⁶² If we extrapolate these numbers to the complaints concerning the duty to refer, we can estimate that there have been over 4,500 complaints since 2010. In consequence, the successful cases based on a violation of the duty to refer would amount to approximately 0,6 per cent.Footnote ⁶³ If we add the fact that the psychological thresholds to go to Karlsruhe – especially in smaller litigations – are extremely high, we can ascertain that constitutional complaints are probably not the most effective way to police the obligation to refer under Article 267(3) TFEU.

Second, the current standard of review boils down to a duty to give reasons.Footnote ⁶⁴ As such, many have criticised that there is no strict scrutiny by the federal constitutional court. Some demand full control of the CILFIT-criteria,Footnote ⁶⁵ others suggest differentiating between references for interpretation, which should be subject to lesser control, and references for validity, which should be subject to stricter scrutiny.Footnote ⁶⁶ A reference under Article 267(1)(b) TFEU serves as substitute for judicial review and compensates for the demanding standing requirement for direct action before the Court of Justice (see Article 263(4) TFEU). Arguably, this argument cannot be extended to the reference for interpretation, which serves primarily the uniform application of EU law. Such a distinction, however, is difficult to sustain. The Court itself has stressed in Köbler – which concerned a reference under Article 267(1)(a) TFEU – that ‘it is, in particular, in order to prevent rights conferred on individuals by Community law from being infringed that . . . a court against whose decisions there is no judicial remedy under national law is required to make a reference to the Court of Justice.’Footnote ⁶⁷ In any case, it should be noted that the Court of Justice has placed an increasing emphasis on the duty to refer. It not only clarified the CILFIT criteria in Consorzio,Footnote ⁶⁸ but also found – for the first time – a violation of Article 267(3) TFEU in an infringement procedure against France.Footnote ⁶⁹ Seen in this light, it seems that the federal constitutional court’s standard lacks behind.

A. Comparative statistics

To start with, German references occupy, in absolute terms, the first place among the Member States. This is hardly surprising considering that Germany is by far the Member State with the largest population. In 2023, German references accounted for approximately 18 per cent (94 out of 518) of preliminary references.Footnote ⁷³ This corresponds exactly to the share of German citizens in the Union’s population. Against this backdrop, it could be argued that German courts are neither very active, nor overly passive when it comes to submitting preliminary references.

Yet, this picture changes once placed in a comparative context. For one, the comparison with the total amount of references reveals that the overall share of German references has steadily declined over the years (Figure 1). Moreover, taking only the absolute references at face value does not offer much insight into whether German judges refer much or little.Footnote ⁷⁵ Instead, many possible factors should be taken into account, such as references relative to population size (Figure 2), references per judges in a certain jurisdiction (Figure 3), or references per incoming cases (Figure 4). Further factors can be government expenditure within the judiciary, a Member State’s share in intra-EU-trade, or the duration of its membership in the EU.Footnote ⁷⁶ All these factors markedly alter the ranking. Specifically, Member States like Austria, Belgium, Bulgaria, Latvia, Luxembourg, or Ireland seem to take the lead, while German judges lack behind and refer below average.

Figure 1. References from Germany in comparison to the total of references per year.Footnote ⁷⁴

Figure 2. References per 500,000 citizens (in 2023) from 2009 to 2023.Footnote ⁷⁷

Figure 3. References per 100 judges (in 2020) from 2009 to 2023.Footnote ⁷⁸

Figure 4. References per 1000 incoming civil and administrative cases before 1st instance courts (in 2020) from 2009 to 2023.Footnote ⁷⁹

These statistics show that Germany occupies place No 15 in references relative to population size (see Figure 2), place No 20 in references relative to the overall number of judges (see Figure 3), and place No 12 in references relative to incoming civil and administrative cases in first instance courts (see Figure 4). It should be added that some have also placed the number of references in relation to the public expenditure in the judiciary. But also here, Germany occupies with place No 22 the back seat.Footnote ⁸⁰

Where then lies the problem? Can we locate the problem geographically? Is there a particular region with low referral rates? That does not seem to be the case. Daniel Kelemen and Tommaso Pavone have traced the emergence of preliminary references geographically. Their maps demonstrate that the distribution of references in Germany is fairly balanced – as is to be expected in a federal state.Footnote ⁸¹ Can we trace this reluctance to a specific branch or tier of courts? Again, the available evidence suggests that this is not the case. The Bundesfinanzhof has been the most referring tribunal among European courts. Since the creation of the preliminary reference mechanism, it has referred 362 cases to the Court of Justice. Among the federal courts, it is followed by the Bundesgerichtshof with 299 references, the Bundesverwaltungsgericht (165 referrals), the Bundessozialgericht (77 referrals) and, finally, the Bundesarbeitsgericht (68 referrals). Lower courts account for 2,106 references and thus for 68 per cent of all German references ever sent to the Court.Footnote ⁸² As Figure 5 demonstrates, this proportion has remained relatively stable over the years.

Figure 5. Share of German references by the type of courts.

A comparative assessment demonstrates that this proportion is close to the European average. Whereas Danish (82 per cent), Spanish (82 per cent), Bulgarian (81,5 per cent) or Belgian (73 per cent) lower courts account for a high share in references, this is less the case for the lower courts in Austria (56 per cent) and Luxembourg (53,5 per cent), and even less so for those in the Netherlands (36 per cent), Finland (28 per cent), Ireland (24 per cent), or Lithuania (23 per cent).Footnote ⁸³ In consequence, we cannot associate the low referral rate with one specific tier of courts. Conversely, these statistics suggest that the German judiciary – taken as a whole – is comparatively reluctant to use the preliminary reference procedure.

Certainly, this kind of quantitative comparison is prone to objections. In particular, one might argue that even if the share of German references is not as large as anticipated, their importance and quality is higher than that of their peers. Yet, notions such as ‘importance’ and ‘quality’ are difficult to define. What makes a case important? The value of the dispute? Hardly. Costa/ENEL concerned an electricity bill of 1,925 Italian Lire, which equals a couple of Euro today.Footnote ⁸⁴ Then its impact on the EU legal order? Or its political implications? Such factors will inevitably involve subjective and controversial assessments. Probably the least controversial yardstick is the Court’s own assessment. According to Article 60 of the CJEU’s Rules of Procedure cases are assigned to chambers of three or four judges or to the Grand Chamber according to the ‘importance of the case.’ Once we look at the statistics, however, German references do not seem to be given more importance than other jurisdictions. Both references from France – an extremely reluctant jurisdiction – as well as from Ireland, Luxembourg, or the Netherlands are allocated to bigger formations than those emerging from Germany (Figure 6).

Figure 6. Allocation of references to formations of the Court from 2009 to 2023.Footnote ⁸⁵

Similar uncertainties will surround the notion of ‘quality.’ What makes a ‘good’ reference? Again, we can only paint with a very broad brush. A possibly ‘bad’ preliminary reference will be an inadmissible one as it does not fulfil its function of ensuring the uniform application of EU law and turns out to be unnecessary. The same might apply, albeit to a lesser extent, to questions which have already been answered by the Court, the reply to which can be clearly deduced from the case law, or the resolution of which does not leave any reasonable doubt. According to Article 99 of the Rules of Procedure, the Court can answer such questions by way of a reasoned order (Figure 7).

Figure 7. Share of inadmissible references/references decided by reasoned order from 2009 to 2023.Footnote ⁸⁶

Although German references are rarely declared inadmissible (around 0.2 per cent) or treated under the simplified procedure (around 0.8 per cent), this figure is close to other jurisdictions, such as Ireland, Luxembourg, or the Netherlands. This indicates that the quality of German references is not much higher than that of many of its peers with higher referral rates.

B. Possible reasons

How then can we explain this relatively low number of references? During the past years scholars have started identifying factors that might incentivise or disincentivise references. As Virginia Passalaqua noted, ‘no single factor exhaustively explains the variation in the reference rates of judges.’ Instead, there is a variety of interacting factors, partly structural and partly subjective which influence judges’ decisions.Footnote ⁸⁷ In this spirit, the general systematisation developed by Broberg and Fenger might be useful in this respect.Footnote ⁸⁸

On a structural or macro level, they contend that several factors might incite the emergence of a suitable case, such as the Member State’s expenditure for the legal system or its share in intra-EU trade. Further, one may add the general constitutional setup of a Member State, namely whether it opted for a strong or rather weak judiciary,Footnote ⁸⁹ or the general willingness of citizens to go to court, including costs or locus standi criteria. With regard to Germany, however, most of these factors – especially the share in intra-EU trade, a strong judiciary, and the accessibility of courts – would rather support a stronger referral rate.

Another structural factor mentioned by Broberg and Fenger is the Member States’ general compliance with EU law. If there is a ‘habit of compliance’, fewer contentious cases might emerge in the first place. Yet, this factor reveals several shortcomings. First, it remains extremely abstract and difficult to determine.Footnote ⁹⁰ Second, it can cut both ways: while less compliance can generate more disputes that give rise to references, better compliance can go hand in hand with greater awareness and thus lead to more references as well. As such, this factor proves to be ambiguous at best. And third, it seems that a ‘habit of compliance’ would rather relate to disputes between public authorities and individual litigants and thus to administrative cases. Nevertheless, the reluctant application of EU law by German courts seems to permeate the entire German judiciary, including civil courts primarily tasked to resolve disputes between private parties.

Finally, one might argue that size matters. As Figures 2–4 demonstrate, Germany is not alone with its reluctant judiciary. Italy, France, Spain, and Poland remain below average as well considering references per population, judges, or cases. Indeed, it is striking that larger Member States are at the lower end of the spectrum, whereas smaller judiciaries take the lead. What might be the reason? Two tentative hypotheses could explain this correlation. On the one hand, a larger population might not necessarily mean more controversies. Instead, there might just be a certain ‘amount’ of controversy that will realistically arise in a certain legal system – irrespective of the size of its population. On the other hand, the larger the judiciary, the more references emerge in absolute terms. As such, it is more likely that the controversy at issue has already been referred to the CJEU. And even though national courts might not be necessarily aware of every judgment from Luxembourg and its potential impact on their own legal order, they will take note of references emerging from their own judiciary. Such preliminary rulings are much easier to digest as they are already embedded in a national procedural context and do not require any translation into the own legal system. Taken together, both hypotheses suggest that at a certain point the curve starts to flatten. If this is true, references will cease to rise in strict proportion to population, judges, or cases. And it might very well be that Germany – as well as Italy, France, or Poland – have crossed this point. In that case, a lower share in references is just an inevitability.

Can this fatalist attitude convince? In the EU context, the first premise seems especially shaky. All human interactions are unique. As such, they can give rise to an unlimited amount of possible legal constellations. Even though there are certainly run-of-the-mill cases, no case will ever be identical. Hence, it seems highly unlikely that there is any limit to controversies that can realistically arise in a given society. This applies especially in the EU, which faces not only the dynamics of society but constitutes a dynamic, constantly changing legal order itself.

Beyond such structural, macro factors, Broberg and Fenger add a behavioural or micro level factor, which concerns the readiness of the individual judge. At this level, they identify additional factors that may drive a judge to refer. Here, one can add a rather fuzzy bundle of factors that relate to the habitus of judges, such as their acceptance to submit themselves to a court beyond their immediate jurisdiction, or the judges’ trust in the Luxembourg court.Footnote ⁹¹ Further, three, sometimes overlapping factors are deemed to be relevant: workload, hierarchy, and knowledge.Footnote ⁹²

Starting with the first, it is self-evident that a high workload can disincentivise referrals, tip the scales in favour of addressing immediate needs, and thus breed ‘a habitus of non-referral.’Footnote ⁹³ As Stone Sweet argued, we ‘assume that national judges strongly prefer to dispose of their cases efficiently, that is, they would like to go home at the end of the day having disposed of more, rather than fewer, work-related problems.’Footnote ⁹⁴ This will apply in particular to judges in career judge systems, where the focus lies on the efficient resolution of cases. While this explains why higher courts, which are usually better equipped and occupied by less cases, refer more than lower courts,Footnote ⁹⁵ there is not much reliable comparative data on the workload of Member State courts. At least the incoming cases per judge demonstrate that German judges do not face a comparatively higher workload than their peers in other Member States (see Figure 4). Therefore, a high workload cannot be interpreted as a particularly German phenomenon.

Second, most authors refer to (the lack of) knowledge, when it comes to referring to the Luxembourg court. At first glance, it may seem questionable that, after 70 years of European integration and a subsequent permeation of EU law into the law school’s curricula, basic knowledge concerning the preliminary reference procedure is an issue. This must especially apply to Germany as a founding Member State. Nevertheless, a study conducted by the European Parliament revealed astonishing deficits among German judges in this respect. 65 per cent of the asked judges and prosecutors replied that they have no or only a minor knowledge of how to refer a case to the Court of Justice.Footnote ⁹⁶ Interestingly, this share was significantly lower than among judiciaries with higher referral rates, such as Austria (44 per cent), Luxembourg (40 per cent), Bulgaria (30 per cent), or Ireland (13 per cent). Whether simple modesty or serious concern, their reply to this study supports the impression that German judges do not feel comfortable when it comes to the preliminary reference procedure.

Third, lower courts may be inclined to pass on the responsibility for referring to higher courts. The German appeal system offers an easy way out. As noted before, in many cases appeals are only admissible if granted by the court whose decision is supposed to be challenged (see Section 1.B). An appeal must be allowed if a case features ‘fundamental significance.’Footnote ⁹⁷ This is the case, when it seems likely that – looking ahead – the court of last instance will face a duty to refer under Article 267(3) TFEU.Footnote ⁹⁸ In consequence, a lower court judge who does not refer is obliged to admit an appeal in case he or she holds a reference to be likely in second or third instance. This option relieves the judge to submit the reference to him or herself. Instead, they can simply grant the appeal – a practice often applied by lower courts.

Finally, the German judiciary seems to have a relatively low level of trust in the Court of Justice. According to a study conducted by Juan Mayoral, Germany scores lowest in terms of trust in the European judiciary compared to other jurisdictions. Only 68.61 per cent of German judges express trust into the Luxembourg court.Footnote ⁹⁹ The factors for such a large distrust are complex. They might result from the German dualist system or the constitutional court’s critical attitude towards EU law. Over thirty years, the latter has set a highly critical tone. Due to the court’s high standing, it seems likely that this scepticism has pervaded the minds and hearts of German judges.

A. Patterns of external resistance

German courts are credited with taking an active part in the judicial construction of Europe by providing constructive impulses for the development of EU law through the preliminary reference procedure. Yet, this is only one side of the story. Under closer scrutiny, a tale of contestation unfolds. Overshadowed by Karlsruhe’s jurisprudence culminating in PSPP, it has almost been forgotten that the first uses of the preliminary reference procedure led to a heated confrontation between ordinary German courts and the Court of Justice. In this spirit, a senior Commission official noted in 1974 that ‘conflicts with various branches of German law have been bound to occur more frequently than elsewhere.’ He then added: ‘Without wishing to dramatize the issue this is, in fact, the main obstacle to an absolute supremacy of Community Law.’Footnote ¹⁰¹

Analysing these early cases of contestation is important. Patterns in the relationship among institutions and their personnel can develop early on and endure even if the wider context has changed. The rocky start in the interaction between the Court of Justice and the German judiciary might thus help explaining the persisting reluctance of many German judges to make use of the preliminary reference procedure. The following section will explore these early uses of this mechanism by contrasting two examples. It will start by focussing on the first open ‘revolt’ of German courts,Footnote ¹⁰² before taking a closer look at a supposed success story, namely the development of EU fundamental rights and the role played by the German judiciary.

Starting with the first example, most of the early German references emerged from tax courts. This was a consequence of one of the Communities’ central aims, which was the creation of a customs union (Article 3(a) of the EEC Treaty).Footnote ¹⁰³ Against this backdrop, a rather unlikely institution became Luxembourg’s central German interlocutor: the Bundesfinanzhof. The other federal courts remained largely in the shadows and engaged relatively little with their European counterpart.Footnote ¹⁰⁴ However, unlike its peers in other Member States, the federal fiscal court did not participate in constructing the new legal order. Instead, its references became a central vehicle to challenge the Court’s early jurisprudence on direct effect and primacy. This led to an – today almost forgotten – period of open jurisdictional conflict between the Court of Justice and the German judiciary.

The first German refence that made a substantive leap in this respect was submitted by the tax court of Saarland.Footnote ¹⁰⁵ In Lütticke it embraced the fairly new doctrine spelled out in Van Gend and asked whether Article 95 of EEC Treaty, which resembles today’s Article 110 TFEU, had direct effect too. If this were the case, it wanted to know whether a specific German turnover tax on imported products would be compatible with this provision. The context slightly differed from Van Gend as the obligations under Article 95 were less specific. Importantly, it included a positive obligation for the Member States to take appropriate measures within a certain time frame. Despite these obstacles, the Court answered in the affirmative.Footnote ¹⁰⁶

By the end of 1967, this led to a flood of administrative appeals and claims for tax refunds in Germany (reportedly 332,000), more than 24,500 of which are said to have resulted in procedures before tax courts.Footnote ¹⁰⁷ Under this pressure, the Bundesfinanzhof referred the question in Molkerei again to the Court, voicing ‘concerns’ and asking in essence whether it actually meant what it stated in Lütticke.Footnote ¹⁰⁸ One can grasp the federal fiscal court’s distress, when stressing that it is not for the German courts ‘to anticipate, by thousands of separate decisions, the action of the legislature. . . or to make good its failure to act’, especially since the infringement procedure confers on the Community the means to compel the States to observe the Treaty.Footnote ¹⁰⁹ The Court nonetheless confirmed its prior interpretation.Footnote ¹¹⁰

In parallel, the federal fiscal court raised objections with regard to primacy. Its first ever reference in Neumann concerned the validity of a regulation imposing a levy on the import of certain products from third countries. While generally adopting an affirmative view on the compatibility of the EEC Treaty with the Basic Law, it questioned whether the Treaty ‘had the effect of transferring to the community the power to legislate’ on matters coming under the fiscal sovereignty of the Member States.Footnote ¹¹¹ In other words, it suggested that the Community institutions had acted ultra-vires. The Court of Justice nonetheless affirmed the validity of said regulation. Faced with this verdict, the federal fiscal court stressed that it is up to the national judge to decide whether and what kind of legal effects a valid provision of Community law produces in the domestic sphere. This includes the question of whether the application of Community law is precluded by provisions of national law, especially in case of conflict with national provisions of ‘greater force.’Footnote ¹¹² This is where national fundamental rights may kick in, which – after a careful assessment – were not infringed in the specific case.

Even more striking was the federal fiscal court’s resistance against the direct effect of directives, which resulted in the first open defiance by German courts. Already in response to a reference by the Munich tax court, the Court of Justice had opened the door towards the direct effect of other acts than regulations.Footnote ¹¹³ This had led to discussions on whether this reasoning could equally apply to directives. In the aftermath of Van Duyn,Footnote ¹¹⁴ Luxembourg acknowledged that individuals could rely, among others, also on the second turnover tax directive.Footnote ¹¹⁵ This had led to severe repercussions in Germany and hundreds of thousands of tax refund claims. Against this backdrop, the Bundesfinanzhof expressly rejected the Court’s jurisprudence as there was ‘no reasonable doubt’ that directives cannot unfold direct effect.Footnote ¹¹⁶ The Court, on the other, reaffirmed its ruling in Becker.Footnote ¹¹⁷ When this reference reached the Court of Justice, the Bundesfinanzhof even sent a copy of its previous decision to Luxembourg, having apparently not the slightest doubt about its position.Footnote ¹¹⁸

A direct confrontation occurred when the Court of Justice was again confronted with a reference by a lower tax court in Kloppenburg. The Court re-emphasised that individuals can, in the absence of any implementation, immediately rely on directives.Footnote ¹¹⁹ The case was then appealed and ended up at the Bundesfinanzhof, which rejected the Court’s interpretation, without re-referring the case to the Court of Justice.Footnote ¹²⁰ Importantly, the federal fiscal court stated that interpretations under Article 177 of the EEC Treaty cannot extend the EEC’s competence by conferring legislative competences to the Community in an area for which it may only issue directives. This episode of open backlash came to an end when the Bundesverfassungsgericht decided in Kloppenburg that the CJEU’s decision was within the borders of sound legal interpretation, which ought to be accepted. Importantly, it stressed that the Bundesfinanzhof, though being a court of last instance and deviating from the Court of Justice’s prior rulings, failed to make a preliminary reference. As such, it infringed the right to a lawful judge under Article 101(1) of the Basic Law.Footnote ¹²¹ However, this EU-friendly jurisprudence by the federal constitutional court did not last for long. Only six years later, it would render the Maastricht judgment, which foreshadowed a long-lasting, highly conflictual period in the relationship between Karlsruhe and Luxembourg.

The second example of early contestation can be found in the conflicts over fundamental rights protection at the Community level. Most scholars seem to agree that the Court’s early fundamental rights jurisprudence can be credited to pushes by German courts. Still, the specifics of this interaction remain controversial. On the one hand, the role of the German constitutional court as a driving force behind these developments has been questioned.Footnote ¹²² On the other hand, the narration shifts between ‘productive dialogue’ and ‘pushback.’Footnote ¹²³

Most accounts start with the judgment in Stauder.Footnote ¹²⁴ This reference concerned a Commission decision permitting the sale of butter at reduced prices for recipients of social assistance.Footnote ¹²⁵ To make such a purchase, the entitled person had to produce a special identification featuring his or her name, which arguably violated human dignity. The administrative court of Stuttgart asked the Court of Justice whether this decision complied with the ‘the general legal principles of Community law in force.’ In its reasoning, it argued that the German system of fundamental rights protection must, at least in part, be guaranteed equally by the Community institutions as an unwritten part of Community law. It closed with an open warning: should the Court not follow this reasoning, then the German courts would intervene (Figure 8).

Figure 8. Excerpt from the ‘Stauder reference.’Footnote ¹²⁶

For some of the Luxembourg judges, like Pierre Pescatore, this reference arose ‘quite unexpectedly.’Footnote ¹²⁷ Yet, it was embedded in a rich German debate.Footnote ¹²⁸ Some had already argued for a more prudent use of the preliminary reference procedure to protect fundamental rights immediately before the Court of Justice.Footnote ¹²⁹ It is out of this thick discursive context, which hardly escaped the attention of the judges at the Stuttgart administrative court, that the reference in Stauder emerged.

While this reference was arguably quite cooperative, the following reference in Internationale Handelsgesellschaft was drafted much more in terms of contestation. The case concerned a regulation that required undertakings to lodge a deposit before receiving a licence for importing or exporting grain. The deposit was forfeited in case the imports or exports were not fully carried out. The administrative court of Frankfurt am Main refered the validity of this regulation to the Court of Justice. In its view, Community regulations had to respect the fundamental rights and principles guaranteed by the German constitution. In the event of contradiction, the primacy of Community law must yield before the principles of the Basic Law.

The Frankfurt court was informed and inspired by a critical domestic discourse.Footnote ¹³⁰ Only three months before the reference, the prominent scholar Hans-Heinrich Rupp had given a speech at the Deutsche Richterakademie, in which he had described the Community as a ‘Herrschaft ohne Grundrechte.’Footnote ¹³¹ In this spirit, the referring court noted that ‘by ratifying the E.E.C. Treaty the Federal Republic has not renounced its rights . . . to protect elementary constitutional rights within the framework of a European Community. It must be assumed that the German legislator agreed to enter the E.E.C. only on condition that . . . essential structural principles of national law are protected in Community law.’ (Figure 9)Footnote ¹³²

Figure 9. Excerpt from the ‘Internationale Handelsgesellschaft reference.’Footnote ¹³³

In the present case, it reached the conclusion that the regulation infringed the freedom of action and disposition, economic liberty and the principle of proportionality guaranteed by Articles 2(1) and 14 of the German Basic Law.

Against this backdrop, the Court of Justice stressed that, to safeguard the uniform application of Community law, the latter cannot be submitted to national fundamental rights standards. Still, the respect for fundamental rights ‘forms an integral part of the general principles of law’ protected under Community law.Footnote ¹³⁴ After a careful examination, the Court affirmed the validity of the regulation. Dissatisfied with the response, the administrative court referred the case under Article 100 of the Basic Law to the federal constitutional court in November 1971.Footnote ¹³⁵ After examining in much detail the two camps in German legal scholarship and the position of its superior court,Footnote ¹³⁶ it stipulated that the primacy of Community law ‘cannot be based on any legal foundation.’ The Court’s position was characterised as being motivated by ‘purely political’ considerations, namely swift integration. Without engaging with the standards developed by the Court of Justice, the Frankfurt court argued that integration would eventually lead to a ‘constitutional and legal vacuum.’ Anticipating the later judgment by the federal constitutional court, it then noted that ‘the national fundamental principles must be observed so long as there is no written constitutional law of the Community.’ Before the referral in this case, Karlsruhe had left the issue open whether it would review Community law against the fundamental rights set out in the German Constitution.Footnote ¹³⁷ In answering the call by the Frankfurt court, it took up the suggested formula and established what would be known as Solange I.Footnote ¹³⁸

What these two examples demonstrate is that the early interactions between the Court of Justice and the German courts through the preliminary reference procedure were highly conflictual. Importantly, these reactions were less driven by the federal constitutional court, which in many cases rather mitigated the backlash, but above all by ordinary courts, such as the federal tax court or the Frankfurt administrative court. These courts made ample use of the preliminary reference procedure to contain and contest the new doctrines of primacy and direct effect. Accordingly, the pattern of resistance did not only emerge long before Karlsruhe’s jurisprudence in Solange I, Maastricht, or Lisbon, but was also intimately tied to the use of the preliminary reference procedure. These early instances of resistance might thus have shaped the understanding of many German judges of the purposes of the preliminary reference procedure – not as a tool of cooperation, but as a means of voicing contestation.

B. Patterns of internal empowerment

At the same time, national courts have embraced the preliminary reference procedure not only as a tool of external contestation, but also as a tool of internal empowerment. Even if this narrative has shaped generations of EU legal scholars,Footnote ¹³⁹ it has been subject to growing criticism. Under the empowerment thesis, judges often appear as power-hungry creatures pursuing their political agenda by judicial means. Court room realities are far removed from this theory. Most judges simply discharge their mandate and resolve the cases pending before them.Footnote ¹⁴⁰ Nonetheless, there are instances, in which the preliminary reference procedure has been used as an empowering tool. Again, it should be stressed that such references can certainly serve legitimate aims (eg the protection of judicial independence or individual rights). Still, the instrumental use of references to broaden the referring court’s powers goes beyond the purposes of this procedure and should be met with a critical eye.

Generally speaking, such an empowerment can unfold in three ways. First, the preliminary reference procedure can allow lower courts to challenge the legislature. This is nothing new. Generally, EU law empowers national courts in two ways, which are both supported by preliminary references. For one, it expands the ordinary courts’ procedural powers. In case of conflict, they can set aside national laws, a power often reserved under national law to apex courts. Referring national legislation to Luxembourg can provide additional legitimacy when disapplying the respective law. This use of the references by German courts has been subject to much scholarly attention.Footnote ¹⁴¹ Far less explored, however, is the fact that EU law also expands the ordinary courts’ substantive powers. National courts are not limited to review the legality of a legislative act, but also the rationality of legislative policy choices. One pertinent example is the so-called consistency requirement.Footnote ¹⁴² In many areas of EU law, national legislation is only proportionate if it is suitable to attain its objective in a ‘consistent and systematic manner.’ As such, the coherence and consistency of policy choices becomes subject to judicial review. An illustrative example is the German state monopoly in gambling. Several courts referred this monopoly to the Court of Justice arguing that it violates the freedom of establishment and the freedom to provide services. Though seeking to combat addiction, the respective act covered only certain games of chance while allowing private operators to exploit other ones with a similar potential of addiction, such as sports betting. According to the referring courts, this policy choice was ‘evidently’ inconsistent.Footnote ¹⁴³ Framed in these terms, the Court of Justice largely followed this assessment.

Second, the preliminary reference procedure may empower the judiciary to challenge the executive. A recent example can be found in references that question the independence of German courts. In 2022, 70 per cent of German judges believed that judicial independence might come under threat, while 67 per cent considered it necessary to reinforce the autonomy of the judiciary.Footnote ¹⁴⁴ Concerns revolve especially around the independence of German courts from the executive: judges at the state level are appointed, appraised, and promoted by the ministry of justice in the respective Land. Further, the executive is responsible for the organisation, administration, and management of the courts, including the allocation of resources. This may create a ‘reward’ or ‘penalty’ system for judicial decision-making.Footnote ¹⁴⁵ To challenge these perceived interferences by the executive and reassert their independence, some courts made a creative use of the preliminary reference procedure. In light of Luxembourg’s pathbreaking judgments on the overhaul of the Polish judiciary, several German judges questioned their own independence thus inviting the Court to declare their references inadmissible. A first reference from the Wiesbaden administrative court was rejected though,Footnote ¹⁴⁶ whereas a second one, which further substantiates these concerns, has been removed from the Court’s register.Footnote ¹⁴⁷

Third, referring a case to Luxembourg can empower ordinary courts to challenge their superior courts. A good example is the bypassing of the federal constitutional court concerning the labour law of churches. The Basic Law guarantees the self-determination of religious groups, which includes their liberty to deviate from general labour law in areas devoted to their mission.Footnote ¹⁴⁸ In this context, the federal labour court had to rule on the dismissal of a divorced doctor at a Catholic hospital, who had been dismissed after remarrying. The Catholic Church considered this as a breach of loyalty. Whereas the federal labour court had sided with the doctor, the federal constitutional court ruled – upon constitutional complaint by the church – in favour of the latter.Footnote ¹⁴⁹ This result was based on a long-standing case law which placed the churches’ right to self-determination at the centre. Especially the question of whether the activity of an employee was linked to the churches’ ethos was largely withdrawn from judicial review. Instead of implementing this verdict though, the federal labour court contested this long-standing, controversial jurisprudence and referred to the Court of Justice.Footnote ¹⁵⁰ By interpreting the Equal Treatment Directive, the Court stated that national judges must verify whether the requirements set up by churches for their employees are proportionate as well as ‘genuine, legitimate and justified in the light of the ethos of the church.’Footnote ¹⁵¹ Applying these criteria, the Bundesarbeitsgericht sided, again, with the doctor. After noting the differences between Karlsruhe and Luxembourg, it briefly assessed whether there were any objections to the primacy of the latter’s ruling, such as ultra vires or constitutional identity concerns, but refrained from referring the case to the Karlsruhe court again.Footnote ¹⁵²

As was noted before, such cases of empowerment do not per se constitute illegitimate uses of the preliminary reference procedure. Nevertheless, they bear the potential to severely alter not only the separation of powers, but also the judicial architecture in Germany. Judicial hierarchies are not an end in themselves: equipped with greater resources, expertise, and authority the function of higher courts is to correct lower ones both in the interests of the rights of individuals and the coherence of the legal order. The empowerment of lower courts through preliminary references can undermine this function. On the one hand, it weakens the corrective function of their superior courts. Theoretically, higher courts can submit preliminary references to reassert their superiority, correct possible misunderstandings, and contain the lower courts within their jurisdiction. Yet, such exercises have had little success so far.Footnote ¹⁵³ On the other hand, the empowerment of lower judges may weaken the authority of higher courts. The bypassing of the constitutional court by the Bundesarbeitsgericht, for instance, caused considerable stir among legal scholars and practitioners.Footnote ¹⁵⁴ Such harsh reactions demonstrate that there is a real fear of deteriorating the federal constitutional court’s authority. Especially in times when constitutional courts all over the world face increasing pressure,Footnote ¹⁵⁵ we should be careful to further weaken the authority of a constitutional court. Such collateral damage is neither in the interest of the EU or the German legal order.

A. The second senate in PSPP: threat of the last word

Starting with Solange I, over Maastricht to Lisbon, the second senate has established a complex jurisprudence that showed its full impact when reviewing the EU’s responses to the financial crisis.Footnote ¹⁵⁸ In the wake of these decisions, it developed three, essentially interlinked review mechanisms: a fundamental rights, ultra vires, and constitutional identity review.Footnote ¹⁵⁹ Eventually, this jurisprudence culminated in PSPP, in which the second senate declared the CJEU’s judgment in Weiss to constitute an ultra vires act.Footnote ¹⁶⁰ The decision provoked an outcry among EU legal scholars and practitioners, accusing the second senate of plunging the Union into a constitutional crisis. Today, PSPP might have become one of the most scrutinised judgments of the last decades. For the purposes of this study, however, it suffices to assess how the second senate handled the preliminary reference procedure.

Initially, the Karlsruhe court underlined that the review mechanisms mentioned before must be applied in the spirit of ‘Europarechtsfreundlichkeit.’Footnote ¹⁶¹ Procedurally, this principle requires that the Court of Justice has an opportunity to interpret the EU law in question. As such, the preliminary reference procedure plays a central role. Substantively, the review mechanisms must be applied with restraint. Importantly, the Court of Justice must be permitted a margin of error.Footnote ¹⁶² Against this backdrop, the specific operation of the preliminary reference procedure in PSPP, but also its (non-)uses in cases such as OMT and Identitätskontrolle I leave much to be desired.

Procedurally, many have argued that PSPP was not a ‘good faith’ attempt to enter into dialogue.Footnote ¹⁶³ On the one hand, the reference and the grounds on which the second senate declared the Court’s response to be ultra vires hardly match. A central consideration of the second senate was the proportionality assessment conducted by the Court of Justice.Footnote ¹⁶⁴ In stark terms, it dismissed the Court’s approach as ‘simply not comprehensible and thus objectively arbitrary.’Footnote ¹⁶⁵ Curiously, the initial reference featured practically no elaboration on issues of proportionality.Footnote ¹⁶⁶ If the second senate wanted the Court of Justice to conduct a detailed, substantive proportionality assessment of the ECB’s actions, it should have made such a desire clear in its reference.

On the other hand, it could have referred a second time to the Court of Justice.Footnote ¹⁶⁷ Even if this second reference would have yielded the same reply, a further referral would have demonstrated the second senate’s willingness to engage in constructive dialogue. Here, the Italian Corte Costituzionale provides an alternative vision for how to structure and resolve a conflict with the Court of Justice. The Corte has been praised for its dialogical attitude throughout the Taricco saga.Footnote ¹⁶⁸ After receiving a first answer from the Court of Justice (on a reference from a lower court), it re-referred the issue. In its second reference, it not only flagged its constitutional identity concerns, but also affirmed the importance of primacy, demonstrated why the CJEU judgment in question was not in harmony with European standards and made proposals for how to remedy the situation. In this sense, the Corte seems to have embraced what Marta Cartabia, its former president, described as a ‘relational style’ of constitutional adjudication.Footnote ¹⁶⁹

Also substantively, the decision in PSPP hardly lives up to the standards of a dialogue in good faith. First, it has been noted that the second senate ignored its previous requirements of restraint and the margin of error granted to its European interlocutor. Along these lines, Julian Scholtes qualified PSPP even as a ‘relational abuse’ of the court’s constitutional identity doctrine.Footnote ¹⁷⁰ Second, such a lack in relationality can also be traced in the decision’s substance. Some have argued that the second senate voiced legitimate concerns. Ana Bobić, for instance, stressed that Karlsruhe sought to remedy the flaws it had over years been warning about – namely to provide more accountability for the Union’s monetary policy.Footnote ¹⁷¹ However, the second senate could have articulated these concerns in a much more relational manner that renders these concerns accessible for the Court of Justice. Instead of referring to the German conception of proportionality, it could have formulated the reference in terms of common Union values, such as the rule of law and democracy enshrined in Article 2 TEU.Footnote ¹⁷²

Finally, it should be stressed that conflicts between the Court of Justice and national courts can be a desirable feature that establishes checks and balances between the different legal orders.Footnote ¹⁷³ The course of European integration demonstrates that this interaction is an important driving force for legal development. Even though this dialogue will inevitably lead to conflicts and open confrontation, it can play an important role in protecting certain constitutional principles. Yet, to untap the ‘deliberative potential’ of constitutional courts in this context, the dialogue with Luxembourg needs to remain constructive.Footnote ¹⁷⁴

B. The first senate in Right to be forgotten II: potential of the first word?

An indication for how a more cooperative dialogue could work in the future was provided by the first senate in its Right to be forgotten II decision. Following the path of other constitutional courts,Footnote ¹⁷⁵ the first senate expressly embraced EU fundamental rights as a yardstick if the case under review is governed by provisions that are fully harmonised under EU law. Hence, it overruled its previous jurisprudence, which distinguished whether EU law leaves discretion for its national implementation or not.Footnote ¹⁷⁶ Whereas the latter was excluded from review, the former was reviewed solely for its compliance with national fundamental rights.Footnote ¹⁷⁷ After this decision, the federal constitutional court could become a guardian of EU fundamental rights.

This has several consequences for the preliminary reference procedure. Internally, it reorganises the judicial architecture. Under the previous case law, ordinary courts were the only ones to review the compatibility of national acts with the Charter and eventually refer these matters to the Court of Justice. The Right to be forgotten II changed this set up. In this context, the first senate expressly acknowledged its duty to refer under Article 267(3) TFEU and contemplated whether it would act – with regard to Charter issues – as court of last instance. This would shift the duty to refer under Article 267(3) TFEU to the constitutional court:

It remains to be seen how Karlsruhe will develop this jurisprudence and whether it intends to monopolise Charter-related referrals thus acting as a sort of gatekeeper.Footnote ¹⁷⁹ Externally, the first senate seeks a much closer, less conflictual cooperation with the Court of Justice. In this spirit, it understands itself as a court of last instance under Article 267(3) TFEU that is obliged to bring such matters before the CJEU and emphasises the latter’s ‘final authority for interpreting EU law.’Footnote ¹⁸⁰

Many have applauded this step. This new approach allows the federal constitutional court to actively engage in the development of EU law.Footnote ¹⁸¹ In this sense, some have interpreted the first senate’s decision as reasserting its authority in matters of fundamental rights.Footnote ¹⁸² Admittedly, national judges cannot instruct the Court of Justice to adopt a specific interpretation. This does not mean, however, that engaging with the Charter constitutes an act of submission. By referring to Luxembourg, national constitutional courts can influence the framing of a case and provide their provisional view on the respective issues. This might nudge the Court of Justice in a certain direction.Footnote ¹⁸³ Power can be found not only in the ‘final’ but also in the ‘first word.’Footnote ¹⁸⁴ This is especially the case for national apex courts. As empirical evidence suggests, the CJEU attaches great importance to references of such peak courts.Footnote ¹⁸⁵

Despite this positive outlook, the realities in the aftermath of the Right to be forgotten II leave us with a mixed picture. Many observers, including judges from the first senate, expected an increasing number of references.Footnote ¹⁸⁶ Especially in the case itself, such a reference would have been possible, some might say even necessary.Footnote ¹⁸⁷ So far, no new references have been submitted by the first senate. As such, the full potential of having the ‘first word’ has not materialised yet. At the same time, the second senate has largely accepted its twin’s jurisprudence.Footnote ¹⁸⁸ While wielding the stick of its review mechanisms, which remain applicable in these cases, it struck a much more conciliatory tune: the guarantee of fundamental rights under the Charter will – usually – not result in a violation of the German constitutional identity.Footnote ¹⁸⁹ The path towards a more cooperative dialogue through the preliminary ruling mechanism has thus been laid – now the constitutional judges need to follow it. As a role model for the entire judiciary, it could thus incite a more positive view on the preliminary reference procedure in the hearts and minds of German judges.