Detecting the Footprints of the Legal Reasoning of the Court of Justice: Inquiry into the Citation Methods

Table of Contents: 1. Simple as it is? – 1.1. Analysis of drafting elements. – 1.2. Interpretative and systematic understanding. – 2. A concise framework to understand the use of precedents by the Luxembourg Court. – 3. Case-study 1: a turning point in the case-law on the independence of judges. – 4. Case-study 2: two close episodes regarding primacy of EU law. – 5. Concluding remarks.

Abstract: In its rulings, the Court of Justice of the EU (CJEU) largely rely on its precedents. Sometimes these references do not seem particularly important, as they might be chosen without a precise method, or are simply aimed at showing that the argument made is rooted in the previous case-law. In other cases, scholars may use citations to trace a point of law back to its precedents and discover non-immediately evident elements. Citations may not attract attention, but learning to use some techniques, apt to disclose the approach of the Luxembourg Court behind the use of references to previous case law, greatly helps, on many occasions, to grasp the way the CJEU looks at its precedents and to catch how it manages the relationships with the different lines of its previous case law. As part of the textual analysis, the aim of the inquiry into citations is, in particular, to discover possible shifts in meaning of certain expressions throughout the chain of citations covering sometimes large time spans, or to highlight undermotivated passages in the legal reasoning contained in CJEU’s judgments, maybe in the context of explicit or implicit overturns. Anyway, citations analysis investigates, at least sometimes, almost casual choices, therefore it should be used carefully and without overestimating its results. It should be combined with other tools to confirm and corroborate the elements emerged, on the basis of an equilibrate approach based on systematic interpretation and a solid theory of precedents, and with a view to avoiding the risks of exasperate originalism applied to case law. Moreover, all elements should be read under the light of historical changes involving sectoral lines of interpretation or even some basic paradigms of Union Law, that can provoke apparent inconsistencies, but that often also show the vitality of the EU legal order.

Keywords: Court of Justice of the EU – legal reasoning – precedents – citations – originalism – systematic interpretation.

1.   Simple as it is?

‘When did the Court of Justice state this for the first time?’: this is a question EU Law scholars often ask themselves. They sometimes also follow the citations[1] the Court of Justice itself often provides as a comfortable means to trace a point of law, contained in a judgment, back to its origin. The reference provided is normally to the latest ruling where the point has been restated and, very often, to ‘the case law cited’: going to that ruling and following again the reference to the other previous precedents makes it possible, after another or some other passages, to reconstruct the complete strain of case-law. Remonter le fil thus becomes a relatively easy exercise and EU Law scholars, well acquainted with the search engine of the CURIA website,[2] should finally be able to discover the ruling where the point of law in question has been made for the first time.

The situations of the different cases in question can hardly be the same and they might come from very different historical, factual and legal backgrounds. This may complicate things a little bit, as the point of law to be taken from previous judgments may have been formulated in a too different context or it might be …a different point of law.

As recent literature has shown, with the use of enquiries applying network analysis to the case-law of the Court of Justice, citations in CJEU’s rulings may openly report an entire strain, but the substantial reference should sometimes be intended to a more limited point of law.[3] On the one side, this requires a critical assessment by the researcher, to identify the effective point of law which constitutes the relevant piece of argumentation used by the Court of Justice. On the other side, the citations may carry also the non-directly relevant points with the effectively useful point, thus making the entire reported sentence as a fixed expression, acquiring normative character as a whole.

Hence, the citations exercise drives scholars through a quite intricated itinerary, and often the Court of Justice may call readers to see a previous decision ‘by analogy’ or ‘to that effect’.[4] Thus, in the course of a reconstruction, there may be a shift in the argumentation – or maybe more than one – that makes unclear whether the Court of Justice is really referring to the last point of law, i.e. to the point made in the last case cited, or at least in the latest case-law cited.[5] Even if the arguments rest the same and where the literal formulation is unchanged, the way the Court of Justice intends such arguments can vary over time. This adds another layer of complexity: even when the Court is making reference to the first case where a certain statement has been made – as it occurs when the Luxembourg Judges are ostensibly willing to stress the importance of the founding building block of a certain strain of case-law – such reference has to be intended not just in its original context, but in the new context and in the way it has been intended and re-construed also by the subsequent case-law …that is maybe not cited (or not directly or not completely).[6] Take primacy: some aspects will be the object of attention in the next pages – where specific references will be provided –, however it can be anticipated that the scope of primacy has clearly broadened over time. Therefore, a reference, for instance, to the seminal judgment in Costa v. Enel[7] is very often not just limited nowadays to the nude idea of the duty to set aside a national provision in contrast with a European one, but also to a series of other aspects, added by the subsequent case-law and/or given by the new meanings the Court is willing to derive from the reasoning in the case at hand.[8]

It can be accepted that an order, a judgment, or an opinion of the Court of Justice enters into the corpus of its jurisprudence and is thus the object of interpretation by the Court itself on new successive occasions. To our limited ends, it should be borne in mind that especially old, fundative, rulings are to be intended in their original meaning, and as enriched by the subsequent case-law. When following the strains of citations, this dynamic has to be borne in mind.

In the light of the elements sketched above, the relevance of an inquiry into citation methods is twofold: on the one side, concrete aspects related to drafting can offer important perceptions of single judgments and lines of case-law; on the other side, citations analysis can offer insights for interpretative and systematic understandings of the jurisprudence of the Court of Justice and of the legal reasoning used. These two ends are pursued in two branches of the analysis (although not hermetically separated), that I will briefly present in the following two sub-sections (1.1 and 1.2): in both of them the interpreter is called upon to make some methodological choices, that will guide him/her in the understanding of the results of the analysis. It has to be stressed that the methodology outlined here relies only on the text of the decisions of the CJEU, with a view to proposing research tools available to all EU scholars. Internal working methods, or the technology used at the CJEU, are not object of the present enquiry, although observation on some well-known general exigencies, related to the management of the ever-increasing workload or to the linguistic regime, will be occasionally made. I will afterwards concisely provide a basic overview of the use of precedents, in general, by the Luxembourg Court, which is especially useful for the aforementioned second branch of the analysis (Section 2). Then, I will present two case-studies (Sections 3 and 4), to test the type of analysis proposed and finally I will offer some tentative conclusions (Sections 4).

1.1.  Analysis of drafting elements

Elements related to drafting per se, of various natures, can be illustrative of important aspects of the logical process followed by the Court of Justice. On top of the simple methods to which I have hinted at the beginning of Section 1, four main sub-categories can be evaluated: elements of a statistical nature; features regarding the quality of the citations; references to other judges’ jurisprudence; the relevance of omissions. The four sub-categories in question should be intended as follows.

a) A statistical research on the number and maybe increase or decrease of citations of a certain decision over time can sometimes be useful. However, results must be interpreted in the light of the methods used by the Luxembourg Court: a sudden decrease in citations may well be due to the mere fact that, from a certain point in time, reference has been made to a subsequent judgment. As a matter of fact, a judgment can be cited just because it is the last ruling restating one, or maybe two or more specific points of law[9]. Mere numbers, including those regarding citations of precedents, are not normally a big preoccupation for law scholars; however, it should be borne in mind that an effort is needed in understanding whether the citations detected are useful or not, while their real or apparent numerousness should be interpreted correctly.

b) Leaving aside numbers, the quality of a certain citation is illustrative: apart from the formulas possibly used by the Court (the already mentioned ‘by analogy’ and ‘to that effect’), it should be for the interpreter to understand whether a certain expression, fixed or almost fixed, is used just as a judicial refrain of a point important for the Luxembourg Judges for some reasons – but that could be reconstrued also on the basis of other arguments – or if it constitutes a unreplaceable piece of the legal reasoning. More in general, the way citations are made, their order and combination can explain how the Court of Justice is building up its legal reasoning.

Moreover, when speaking about citations, one could also draw a distinction between cases in which there is a reference to a previous judgment, with the use of exactly the same words of the passage to which reference is made, so that here we should talk about (self-)quotations by the Court of Justice, and cases where the wording is changed more or less profoundly (and with or without the use of formulas about analogy or pseudo-analogy)[10].

In other cases, the Court itself reinforces citations, by stressing that a certain point of law comes from the previous ‘case-law of the Court’. More strongly, in the IS case – rested unique in this respect, to my knowledge –, the Court has reported a sentence from Van Gend en Loos, with the exact original wording and in quotation marks: the Court does not use quotation marks when reporting its own words, but the use of this technique here clearly shows the stress put on the stableness of certain orientations (please see following Section) and a kind of pedagogical intent towards the referring judge (and/or other national authorities) [11].

c) Furthermore, although the CJEU is seen as quite reluctant to establish links with the jurisprudence of other judges, this is done in a certain number of cases. This may be due to the formal competence in a certain field and/or the authoritative character of the rulings of such judges. Interrelations of lines of reasoning can take many different forms. For instance, when responding to the arguments of the parties, the CJEU may give its own interpretation of other judges’ precedents.[12]

d) Finally, inquiry into citations can be more complicated if one wants to look at omissions, that could be as meaningful as citations. Demonstrating that an omission does not come out of simple inaccuracy, but is the result of a choice is an uneasy task: however, scholars can consider the choice not to cite a certain previous judgment at least as a working hypothesis, which, with other concurring elements, may offer possible interpretative paths.

Each time drafting elements of the types referred are identified, they should be correctly understood, as they clearly come out of techniques accurately used and checked internally in Luxembourg (the common use of citing the last judgment, or the fixed expressions referred above hinting at analogical or pseudo-analogical processes are just the most evident signs of the existence of house rules at the Kirchberg Court). Besides, the mere fact that the Court departs from the usual techniques in a given situation may be a sign that the Judges are willing to say something different in such a case. At the same time, such elements should not be overestimated, where their use appears accidental, especially where one wants to cover a wide span of time in a jurisprudential research – as house rules and habits have changed over time –, and due to the fact that various hands touch the texts at different stages, i.e. law clerks (référendaires, in the French jargon of the Court), Judges and readers of judgments (lecteurs d’arrêt, whose job is very important, as they should ensure consistency in the language used and in drafting techniques, as well).

The existence of the referred type of elements could also be the starting point for the functioning of artificial intelligence (AI) tools. I will not deal with this issue; however, the use of the referred techniques, habits and interpretative understandings should take into account the foreseeable use of AI tools by practitioners and researchers (for instance with a view to providing the most useful results and/or to offer interpretative hypotheses concerning the use of citations and/or specific arguments in a certain judgment, or in a line of case-law). Giving the right input and offering the final interpretation should nevertheless remain in the human hands of scholars or practitioners.

1.2.  Interpretative and systematic understanding

Each time a segment of a legal reasoning is borrowed from a previous judgment, the Court of justice is giving a clear message as to its reliance on the corpus, or, as one may say, heritage, of its previous jurisprudence and on the auctoritas – i.e. authoritativeness – that should be recognized to its precedents. As already pointed out and extremizing here, for the sake of clarity, two opposite situations that often occur, there might be a clear departure from the original meaning, due to a physiological evolution of its understanding, or a disguised recourse to a new interpretation or to a new method. In the first case, it will be interesting to catch how the Court sees the evolution of its own judicial acquis, while, in the second, scholars could concentrate on the implicit line of arguments and speculate on the possible reasons for not having exposed it openly.

In both cases, and especially in the second one, the fact that the Court relies on itself is very telling: all in all, the Court wants to reassure the readers that its jurisprudence tends to be stable. Short of a formal stare decisis rule in EU Law, the Luxembourg Judges are clearly willing to provide orientations that should be always followed at all levels, especially within national legal orders, while, at the same time, keeping a not too little room for manoeuvre in the field of interpretation. Each time this latitude may vary depending on many different features of the concrete situation, however, a clear message emerges on the existence of a stable judicial nucleus constituting the fabric on which the Luxembourg judicial machinery builds its case-law. There are theoretical and practical outcomes of this attitude of the Court of Justice, on which I shall come back briefly in the concluding remarks. However, such attitude must be regarded also in the light of two other aspects, that should be recalled already at this stage.

First, explicit overturns of previous judicial orientations are very rare[13]: contradicting openly what the Court itself said in the past is disliked at the Kirchberg. This attitude stems as a corollary of the idea that the past orientations should rest as stable as possible. Therefore, the eyes of EU law scholars should stay wide open on hidden overturns: while it should not be the job of the Court of Justice to conduct covert affairs, it is true that sometimes adjustments, if not radical changes, have been made without announcements. As a consequence, precedents might have been invoked, but sometimes without clearly explaining the reasons why they have been partially or wholly discarded, so that the changes in question have rested undermotivated. Open overturns can all the more be problematic, especially when the arguments used by the Court are not enough articulated and well-founded to lead to change orientation[14].

Second, the well-known interpretation provided by the Court of Justice on the duty incumbent on jurisdictions of last resort to refer a preliminary ruling according to Article 267 TFEU is clearly founded on the premise that an existent judicial orientation should be always considered good law until the Court of Justice itself – possibly solicited by a national court willing to provoke an overturn and to provide good arguments to do so – revises it. In fact, as widely known, a national jurisdiction of last resort can avoid referring a preliminary question, where the response comes out of precedents of the Court of justice addressing an identical question, or an analogous one[15].

Formally speaking, the Court of Justice cannot be considered a jurisdictional organ acting under a common law system. Nevertheless, from time to time the question arises in the literature as to if and to what extent the Court of Justice adjudication scheme imitate that of a ‘common law’ court.[16] I will not discuss the theme here – while a few observations will be made in the following and last Sections –, however I would like to stress that the Court of Justice shows great care of its own precedents and it builds almost always upon the idea that each decision can be used as a block for future ones. This is probably done to produce stableness and certainty, with a view to ensuring clarity and predictability and, therefore, going back to the precedents is useful to reconstruct the full meaning of the arguments used each time by the CJEU.

Nonetheless, a too large reliance on the judicial acquis of the Court, presents also risks on the substantial plane, which have essentially two sides. First, when the legal reasoning remains shrunk in a series of references too deferent to previous cases, the rulings may sometimes suffer of a not sufficiently clear motivation. Second, when the story of the blocks used is not completely coherent, the understanding of the final motivation rests, again, difficult.

2.   A concise framework to understand the use of precedents by the Luxembourg Court

In his study on ‘Precedents and Case-based Reasoning’ by the Luxembourg Court, the reference book for these matters to date, Marc Jacob affirms that ‘the use of precedents by the ECJ is far-reaching but rarely, if ever, explicated in detail as an intricate legal technique in its own right’[17]. Consequently, and as already stated, it is quite complicated to understand how precedents are really used at the Kirchberg and if a comprehensive – although maybe not always coherent and constant over time – precedent doctrine is adopted. By contrast, a formal stare decisis rule does not operate in the EU legal order. While this study is dedicated to the illustration of a reconstruction technique, as part of the textual analysis, the ultimate aim is clearly to help EU scholars to reconstruct the use of precedents, therefore a few words will be devoted to recall a general framework on that issue.

In general terms, using precedents implies different phases[18], that can be sketched as follows. First, a previous case has to be identified and its meaning, i.e. the meaning of its legal reasoning, should be determined. This process implies the consideration of the particular circumstances of the case, including its factual context. Second, once the significance of its legal reasoning has been assessed, a point of law should emerge. Third, the point of law in question will exercise some kind of constraining effect on the case at hand, or, under certain conditions, the precedent should be considered not relevant or overruled.

Enabling a general rule – or a principle – to produce effects also in cases subsequent to that in which a judicial decision has originally expressed/recognized it, by acknowledging its validity and significance beyond the boundaries of that original case is the essence of precedent-based reasoning. Both the identification of the links of the decision to the specificities of the case and the logical process that enables to infer a general rule or principle from a concrete case are influenced by pertinent principles existent in the legal order and by the methods used by the competent courts. Also the kind and extent of the constraining effect on subsequent cases, as well as the identification of the conditions under which precedents would be not relevant or overruled will depend on the general framework established under the considered legal order.

A rule or principle going beyond the boundaries of the decision in which it was expressed is not necessarily a rule or principle interpreted in an extensive way. Further application means application of that rule or principle to another and new similar case, by virtue of analogy or other logical process, that could either expand or restrict the scope or extent of the norm in question. It should be borne in mind that common law systems, where precedent-based reasoning is acquired as a basic judicial rule, rely more on trust than civil law system. As Michal Bobek puts it,

‘[t]he common law tradition (…) is arguably based more on trust and coordination through precedents than on extensive hierarchical review. Each tradition thus creates a different type of authority and officialdom: coordinated on the one hand and hierarchical on the other’[19].

The Author continues by holding that, in this form of ‘coordination’ there will be less need for hierarchical and rigid control over judicial decisions, as, one may argue, the solidity of judicial-making should be rooted in the faithful application of the precedent-based argumentation, that would lead to adjudicating a case. The complex flow of judicial decisions identifies paths of evolution: traditional justifications for the use of precedents include the need to ensure certainty, equality, judicial efficiency, impartial justice instead of arbitrary decision making and, ultimately,

‘(…) the concept of stare decisis reflects very basic notions about the proper function of judges in the lawmaking process’[20].

On top of trust, tradition has a great importance for instance in the American system, intended as a kind of interpretation of the Supreme Court that

‘(…) signals presumptive influence of political and cultural practices of substantial duration for informing constitutional meaning’[21].

This appears different from precedent-based reasoning, but it seems not used in Luxembourg, in particular in explicit terms. The Court of Justice prefers its written precedents, while tradition may be referred to the way it interprets them or specific points of law, but the CJEU seems reluctant to openly resort to tradition.

Let us now look at the attitude of the Court of Justice more broadly. It should be accepted that the primary concern of the Court of Justice is reinforcing its legitimacy by referring to precedents, relying, in a way, to the story of its long standing authoritativeness and to the final acceptance of it by the Member States[22]: also when its authority is openly contested, a fortunate story is still very telling (and hopefully persuasive) and worth relying on. In this connection, one may argue that the Court of justice is firstly saying its precedents – read in the way it is using them at present, i.e. in particular when speaking to national authorities and namely judges – are binding more on national courts, especially Supreme and Constitutional courts, than on itself. Referring to precedents is thus a way to recall that the old story of EU law has already made its own way and must be still accepted to keep the whole edifice stay in place[23]. This is especially true for the references to the big classics of EU law[24].

While the CJEU’s use of precedents is very linked to the case specificities, it emerges the strong will of the Court of justice to show that the building blocks of its legal reasoning can fit into cases sometimes also very different from those where they had been used for the first time[25]. The underlying and implicit assumption is that there is an inherent consistency in the law of the EU, that enables to use such building blocks in different situations and that tout se tient, with connections between basic concepts re-discovered many times, especially when dealing with the very foundations of EU law[26] …and with a little effort to be done to show that each citation really suits well …sometimes just needing to say: ‘see, to that effect’. I believe this tendency is the outcome of the strong will to show that legal reconstructions and principles, methods of interpretations and, ultimately, the identity of the legal order rest the same in many different aspects. While this may feed also the scientific identity of EU law, a too large use of the same line of argumentation in sparse and diverse strains of case-law could also water down the efficacy of the citation methods in the long run[27], or hide shifts in argumentation related to certain principles or rules of EU law[28], and, what is more worrying, could weaken the persuasiveness of the specific legal reasoning used on each occasion. The possible emergence of this risk adds another good reason to take into account citations when analyzing the case-law of the Court of Justice.

Furthermore, the repetitive use of statements that tend to become fixed with a sometimes confused reference to a strain of case-law (where only one case is cited, maybe chosen in a casual way) not only reveals insistence on stableness of jurisprudential orientations, but hides the analysis of the specificity of each case, which is, as seen, essential in the logical process of applying precedents. This makes also less clear if the Luxembourg Court is relying also on other elements, that stem from cultural or political options, or from tradition (whatever conception or tradition one may adopt). In this vein, citations analysis could open interpretative avenues that, with the help also of other methods, may shed some light on the arguments effectively followed at the Kirchberg.

3.   Case study 1: a turning point in the case-law on the independence of judges

Different lines of case law, openly considered by the Court of Justice or not, may at some point overlap or influence each other, and at the end, one of the two may fully or partially absorb the other. Something like this envisaged situation, coupled with one of the rare explicit overturns, seemed to occur in a particular turning point of the case law on the independence of judges, that is the object of the first case study.

The issue of independence of judges, as framed under the judgment in a line of case law inaugurated by Vaassen-Goebbels[29], followed by other judgments, like Pretore di Salò[30], Danfoss[31], Almelo[32], Job Centre[33], Dorsh Consult[34] and others, regarded the functioning of the Article 267 TFEU mechanism. In this case-law, it is required that, to refer preliminary questions to the Court of Justice, a national organ has to fulfil some conditions under which an organ can be qualified as a jurisdictional one in EU Law: it must be established by the law (and not by the parties to the disputes), be permanent, settle disputes with decisions binding upon parties, by applying normative rules, and in an adversarial procedure: according to a common expression, used also in the English version of the relevant judgments, it is said that the procedure should be inter partes, even if this last requirement is not considered per se ‘absolute’. Moreover, such organs must be independent, hence independence, in this context, has been treated for long as a question simply regarding the possibility to refer a preliminary question to the CJEU and thus it constituted an admissibility criterion.

A first interesting point regards the fact that independence is a requirement added by the case law at a second stage, as it results from the statement in Almelo whereby ‘[t]he Court has extended those criteria, pointing out in particular the need for the court or tribunal in question to be independent’[35]. In this passage, reference is made to Pretore di Salò, where the relevant affirmation sounds like this: the Luxembourg Court can reply if the reference comes from ‘a court or tribunal which has acted in the general framework of its task of judging, independently and in accordance with the law’ (emphasis added). The adverb ‘independently’ thus turned out to be a heavy one. The reference in question is also extended to Fratelli Pardini[36], where the sentence just quoted is reported literally, and to Corbiau[37], where the formulation is completely different and the word ‘independence’ (or others with the same root) is not used; it is, however, stated that a court or tribunal must be a ‘third party’: an organ, which is fully integrated in the structure of a public administration, and that, in case its decision had to be challenged, would be a counterparty of the plaintiff before the competent judge, cannot be considered as a ‘third party’[38].

Let us now come to the story of our turning point.

In Gabalfrisa,[39] a judgment delivered in 2000, the Tribunal Económico-Administrativo, existing under a Spanish law, was considered a jurisdiction within the meaning of Article 267 TFEU. The Court of Justice started by citing Dorsch Consult ‘and the case-law cited therein’,[40] then examined at some length the features of the Tribunal in question, as laid down in Spanish law: it ascertained its permanent nature,[41] the compulsory character of its jurisdiction[42]  and the binding force of its decisions.[43] As to the inter partes character, the Court recalled that in Dorsch Consult it considered such requirement not absolute, observing, nonetheless, that the procedure before the Spanish Tribunal Económico-Administrativo met such requirement.[44] Then, the Court went on recognizing the Tribunal was invested with the task of applying ‘rules of law’.[45] Finally, as regards independence, it is worth reporting the relevant passage:

‘Finally, it is important to note that Article 90 of Law No 230/1963 ensures a separation of functions between, on the one hand, the departments of the tax authority responsible for management, clearance and recovery and, on the other hand, the Tribunales Económico-Administrativos which rule on complaints lodged against the decisions of those departments without receiving any instruction from the tax authority.

Such safeguards give the Tribunales Económico-Administrativos, unlike the Directeur des Contributions Directes et des Accises (head of the Direct Taxes and Excise Duties Directorate) in question in Case C-24/92 Corbiau v Administration des Contributions [1993] ECR I-1277, paragraphs 15 and 16, the character of a third party in relation to the departments which adopted the decision forming the subject-matter of the complaint and the independence necessary for them to be regarded as courts or tribunals for the purposes of Article 177 of the Treaty.’[46]

Thus, by marking a difference between the Spanish organs in question and those considered in Corbiau, the CJEU acknowledged the separation of functions between the Tribunal Económico-Administrativo and the structure whose decisions can be challenged before it.

The story could be considered quite conventional up until this point, leaving aside any physiological critical attitude towards decisions of the Court of Justice that, like those of any other judicial organ, can be analyzed and criticized. What it matters here is that the Court of Justice, twenty years later, changed its mind and, in the Banco de Santander ruling,[47] made an explicit overruling exactly on the point of the qualification of the Tribunal Económico-Administrativo.

It could not be said that the old case law on independence within the meaning of Article 267 TFEU was abandoned, but it seems that decisive arguments were taken from a more recent line of case-law which was starting to construe the idea of independence of the judiciary as a building block of the rule of law, one of the values of the EU under Article 2 TEU.

Let us follow the citations.

First, the classical case law centred on Article 267 TFEU, with the well-known requirements for an organ to qualify as a national jurisdiction capable to refer preliminary questions, was recalled and it is significant that the Court of Justice mentioned Vaassen-Goebbels, Syfait[48], a judgment of 2005, Margarit Panicello, delivered in 2017,[49] ‘and the case law cited’.[50] The reference is a quite conventional one, as regards the recalling of the case-law cited, but it is not limited to the last judgment; it is interesting that three judgments come from three different periods of time and this has not been done by chance and the Court significantly started with the first, that initiated the strand in question. As a hypothesis, it can be argued that the drafters wanted to stress the reference to the entire developing line, starting with its roots, with a view to underlining a certain sense of faithfulness to a well-established orientation. This seems tantamount to a clarification that the overruling would regard just the solution regarding the Spanish organ in the case at hand, not the complex of the ordinary criteria, which still stay in place.

When the CJEU went on in its reasoning, it was immediately stated that all the criteria set out in the case-law in question were fulfilled, with the exception of independence, which then became the focus of the subsequent line of reasoning: after having recalled Gabalfrisa, the ruling about to be overruled, the Court invoked three times Associação Sindical dos Juízes Portugueses (ASJP)[51]. The first time, which appears the most meaningful, regarded the idea that independence is a strict requisite for being entitled to refer a preliminary question and it must be said that the classical case law was also recalled again (with reference to the previous passage) [52] …but ASJP was not a case in which there were doubts on admissibility![53] In that judgment, independence was strictly seen under the light of the idea that, even on a foreseeable occasion, it must be a distinguishing feature of an organ which enters in dialogue with the Court of Justice. It must be paid attention to this point: the issue – it should be restated – did not regard admissibility, as there was not a concrete question with reference to which the exercise of effective jurisdictional activity could be evaluated.[54] Therefore, this kind of operational independence could be verified only where the organ in question had to refer a preliminary question to the CJEU. By contrast, ASJP read independence as a necessary characteristic of a jurisdictional organ that may be called upon to apply EU law and, where deemed opportune, to refer to the CJEU. The problem was thus treated as a systemic one, in connection with the general duty of Member States, laid down in Article 19 TEU, to provide judicial remedies for the protection of rights originating in EU law.

Under the Vaassen-Goebbels line of case-law, the Court of Justice evaluation is limited to the question whether that particular organ, in the performance of the specific powers involved in the case at hand, can be qualified as a jurisdictional organ. If the answer is in the negative, there are normally no dramatic implications, but the declaration of inadmissibility of the preliminary reference. This would normally imply that, at a later stage, the issue can pass through a proper judicial avenue before a jurisdictional organ within the meaning of Article 267 TFEU, which may, at that point, decide to make a new preliminary reference to the CJEU.[55]

By invoking a different line of case law, the Court of Justice was taking into account (also) the acquisitions of its own jurisprudence on the evaluation of the very existence, at some point, of a proper judicial organ, before which appropriate judicial remedies can be activated with the guarantees required by Article 47 of the Charter of fundamental rights.

The Court of Justice then turned to analyze whether the Spanish Tribunal was independent under the two dimensions identified by the European case law: the external one, requiring that the organ in question be autonomous and immune from hierarchical coercion or influence of any other body, and the internal one, pertaining to the equality of the parties before the organ in question and its impartiality, its sole interest being the application of the law. It is interesting that judgments from the classical Article 267 TFEU requirements case law were cited with reference to both dimensions,[56] but, as regards the first, they were merged with cases taken from the new strain concerning the rule of law. In the context of the external aspect of independence, and as regards in particular irremovability of judges, the CJEU recalled the already mentioned ASJP[57] and Commission v Poland (Independence of the Supreme Court), a landmark ruling in the ambit of the line of jurisprudence on the rule of law.[58]

It is worth recalling that the passage of Commission v Poland to which the Court made reference in Banco de Santander was an original one, i.e., there were no other judgments to which the CJEU again readdresses the reader. However, by enlarging the reference to the preceding and subsequent paragraphs to the one cited in Commission v Poland, there are citations of Minister for Justice and Equality (Deficiencies in the system of justice) (LM),[59] which is a case on the European arrest warrant, defining when systemic deficiencies in the judiciary of the issuing Member State provide grounds to refuse surrender. The problem emerged with the Polish Judiciary also in LM and it was one of the cases where the forced retirement of some judges was at stake, thus irremovability turned out to be a crucial requisite in that case. If one looks at the judgments to which LM, on its turn, refers back, one may, not surprisingly, find ASJP and Wilson, a case of the classical line of case law, cited also in Banco de Santander.

Independence and namely irremovability has been driven through a journey to classical situations, where the problem was not particularly sensible, and to others in which it assumed a much more delicate role, its lacking implying systemic deficiencies in the Judiciary of a Member State, that could lead to the virtual (…or effective) exclusion of or more of its organs from the circuit of the dialogue with the Court of Justice and/or with the Judiciaries of other Member States.

Turning back to Banco de Santander, in a completely different situation, the Spanish Tribunal was deemed to be too much exposed to external influence, namely of the national executive power, and irremovability of its Members and of its President was considered as not sufficiently guaranteed.[60] After having carried out an analysis of the pertinent national legislation, the Court of Justice marks the difference between the situation of the Spanish organ at hand with that of two other cases, where irremovability was the object of attention, but it was ascertained that the guarantees where sufficient: Consorci Sanitari del Maresme,[61] where irremovability was ensured to members of the body in question up to the end of the term of office, with limited exceptions for grounds provided by the law, and MT Højgaard and Züblin,[62] where, in a body composed of experts and other holding the status of magistrates, only the latter enjoyed constitutional protection, who constituted in any the case the majority. Interesting as they may be for the reconstruction of the requisite of irremovability – because they clarify that such requirement should evidently not to be intended as absolute – these references are quite meaningful, from the perspective adopted in these pages, for another reason. Consorci Sanitari del Maresme and MT Højgaard and Züblin have been delivered respectively in 2015 and 2016 – so quite a lot of years after Gabalfrisa – and they are both cases in which the question of independence was raised in the sole and only context of admissibility, like in the traditional vein of case law on Article 267 TFEU. It cannot be denied that these examples have been useful for the Court of justice to find a foundation for the concrete assessment of the requisite of irremovability in the new analysis of the Tribunal Económico-Administrativo carried out in Banco de Santander. However, the references of the Court of Justice were up to date to the last developments in the line regarding admissibility, while, and probably most significantly, such line was merged with the acquisitions coming from the rule of law jurisprudence[63].

In this connection, it should be noted that, when starting in Banco de Santander the reasoning justifying the overruling of Gabalfrisa, the CJEU stated that ‘as the Commission also submitted in its written observations, [the] considerations [exposed in Gabalfrisa] must be re-examined notably in the light of the most recent case-law of the Court concerning, in particular, the criterion of independence which any national body must meet in order to be categorised as a ‘court or tribunal’ for the purposes of Article 267 TFEU.’[64] However, this statement is immediately followed by the first quotation of ASJP, therefore, and in the light of the foregoing analysis, the ‘categorisation as court or tribunal’ is not read under the sole light of questions of admissibility.

The overruling of Gabalfrisa was not accurately motivated: it was just said that in that judgment the separation between the Tribunal and the administrative structures to which it was in some way connected had been judged sufficient (and this is the only occasion on which reference is made to that previous ruling), however, upon request of the Commission, such conclusion had to be re-examined, as just reported.[65] What follows is a new assessment under the new requisites, without clarifying where and to what extent the old requisites – or the old way to intend the classical requisites – had to be deemed not good law anymore. By comparing the two legal reasonings, it emerges that the regime of removal of members of the Tribunal, which could come out of a decision of the Minister for Economic Affairs and Finance, was just mentioned but not assessed in the light of independence of the organ in Gabalfrisa.[66]

The crossing of references of the two lines of case-law, reported here in some detail to show how the citation analysis can be of help in understanding some passages, can lead at least to a conclusion, that can be summarized as follows. The case law on requisites of independence for the categorization of an organ as ‘court or tribunal’ within the meaning of article 267 TFEU has evolved over time and, at the time of Banco de Santander, there were already some interesting precedents, regarding for instance, as seen, irremovability. Moreover, in that judgment, that stream of case law overlapped with another, coming from a different origin, although dealing also with independence of national organs that should be able (and free) to refer preliminary questions to the Court of Justice. Thus, the idea of independence was enriched with the acquisitions coming from the line of case law regarding the respect of the rule of law and the need (not just the possibility with reference to a specific organ) that national organs able to come in touch with the CJEU are in place and, therefore, legal actions for the protection of rights stemming from the EU legal order can be brought before them.[67]

From Banco de Santander onwards, the dividing line between the two streams of case law on independence has become increasingly blurred, with issues of admissibility of preliminary questions often becoming also more substantial and highly sensitive.

This merger has been the object of criticism on many occasions, as it has been submitted that keeping the assessment of independence separate for the two purposes – admissibility on one side, and assessment on the context of the respect of the rule of law (with systemic implications) on the other – would have ensured more clarity and a more appropriate adjudication in the different contexts.

4.   Case-study 2: episodes regarding primacy of EU law

As a second case-study, I will analyze two judgments where the CJEU refers, also in the operative part of judgments, directly to primacy per se, alone or in combination with specific provisions of EU law. Primacy, as widely known, is a complex and, recently, increasingly controversial legal concept. Different conceptions of primacy have been proposed, that will not be discussed here. However citations will help to highlight some useful elements to understand how the CJEU’s conception of primacy has changed over time.

The two cases in question are Euro Box[68] and RS. The issue of a clash of the national provisions in question with primacy per se was raised already in the proceedings before national judges.[69] In the second one, the problem of the freedom of national judges to adjudicate on matters of EU law, already decided by their Constitutional court was at stake: common judges had in fact to follow the decisions of the Constitutional court as regards compatibility of national norms with EU law, even where they had to be in contrast with orientations expressed by the CJEU. Moreover, referring a preliminary question to Luxembourg, in this kind of situations, constituted a disciplinary offence for the judges who had to decide so.

After an ample reconstruction of the essential features of the EU legal order, the Court of Justice stated the following:

‘50  Following the entry into force of the Treaty of Lisbon, the Court has consistently confirmed the earlier case-law on the principle of the primacy of EU law, a principle which requires all Member State bodies to give full effect to the various EU provisions, and the law of the Member States may not undermine the effect accorded to those various provisions in the territory of those States (judgment of 21 December 2021, Euro Box Promotion and Others, C‑357/19, C‑379/19, C‑547/19, C‑811/19 and C‑840/19, EU:C:2021:1034, paragraph 250 and the case-law cited).

51   It follows from that case-law that, by virtue of the principle of the primacy of EU law, a Member State’s reliance on rules of national law, even of a constitutional order, cannot be allowed to undermine the unity and effectiveness of EU law. In accordance with settled case-law, the effects of the principle of the primacy of EU law are binding on all the bodies of a Member State, without, inter alia, provisions of domestic law, including constitutional provisions, being able to prevent that (judgments of 17 December 1970, Internationale Handelsgesellschaft, 11/70, EU:C:1970:114, paragraph 3, and of 21 December 2021, Euro Box Promotion and Others, C‑357/19, C‑379/19, C‑547/19, C‑811/19 and C‑840/19, EU:C:2021:1034, paragraph 251).

52   Thus, given that it has exclusive jurisdiction to give the definitive interpretation of EU law (see, to that effect, judgment of 2 September 2021, Republic of Moldova, C‑741/19, EU:C:2021:655, paragraph 45), it is for the Court, in the exercise of that jurisdiction, to clarify the scope of the principle of the primacy of EU law in the light of the relevant provisions of that law, with the result that that scope cannot turn on the interpretation of provisions of national law or on the interpretation of provisions of EU law by a national court which is at odds with that of the Court (see, to that effect, judgment of 21 December 2021, Euro Box Promotion and Others, C‑357/19, C‑379/19, C‑547/19, C‑811/19 and C‑840/19, EU:C:2021:1034, paragraph 254).

53   In that regard, it must, inter alia, be recalled that, in accordance with the principle of the primacy of EU law, the national court called upon within the exercise of its jurisdiction to apply provisions of EU law is under a duty, where it is unable to interpret national law in compliance with the requirements of EU law, to give full effect to the requirements of EU law in the dispute brought before it, by disapplying, as required, of its own motion, any national rule or practice, even if adopted subsequently, that is contrary to a provision of EU law with direct effect, without it having to request or await the prior setting aside of that national rule or practice by legislative or other constitutional means (see, to that effect, judgments of 9 March 1978, Simmenthal, 106/77, EU:C:1978:49, paragraph 24; of 24 June 2019, Popławski, C‑573/17, EU:C:2019:530, paragraphs 61 and 62; and of 21 December 2021, Euro Box Promotion and Others, C‑357/19, C‑379/19, C‑547/19, C‑811/19 and C‑840/19, EU:C:2021:1034, paragraph 252).’[70]

Four elements regarding primacy can be extracted from each of the reported paragraphs. First, in paragraph 50, primacy is defined as ‘a principle which requires all Member State bodies to give full effect to the various EU provisions’ and, consequently, national norms shall not constitute an obstacle to the full effectiveness of EU law. Under this light, primacy is particularly linked to the duty of sincere cooperation (see especially Article 4(2) TEU), while the judicial source we discover through citation is Euro Box, where an identical statement is used, ‘and the case-law cited’.[71] By way of the chain of citations, with which the reader will now be well acquainted, the reference goes ultimately and, unsurprisingly, to Costa v ENEL.[72]

In the case at hand, the formulations at hand are highly standardized and repeated identically in many judgments. What may be interesting is that, following the chain, one may find judgments in which cases which were recent at the time and that were considered significant because they added something new, or re-stated something the Court of Justice considered important, are cited many times in the same ruling with reference to similar issues.[73]

Second, in paragraph 51 it is specified, in essence, that even national constitutional provisions shall not constitute an obstacle to the full effectiveness of EU law. As to citations, after recalling again Euro Box, the Court refers back to Internationale Handelsgesellschaft, but with a formulation different from the original, which, albeit clear, was slightly less direct and confrontational: the new one seems to have been appeared for the first time in Winner Wetten.[74]

Third, para 52 clarifies that, as it stems from the interpretation monopoly of EU law held by the CJEU, it is for the same Court to interpret ultimately the principle of primacy of EU law. Two separate citations call respectively into question Republic of Moldova, with a focus on autonomy,[75] and again, with overlaps in the respective strains, Euro Box, stressing the binding nature of the interpretation by the Court. It is interesting that, in this second vein, the CJEU reminds its role in setting orientations that should guide also national courts of last instance, within the limits lastly restated in Consorzio Italian Management, which has been already recalled also in these pages.[76]

Fourth, the classical dimension of primacy regulating contrasts between EU and national rules – a conflict rule – is reconstructed in paragraph 53. The duty to set aside national norms in contrast with EU provisions is put under the umbrella, unsurprisingly, of Simmenthal and of Popławski, a relatively recent case, where, finally, the duty in question was explicitly limited to EU norms with direct effect.[77] Euro Box, as the most recent and important case related to very similar issues, was recalled again.

Let us turn briefly to the latter now. Here the Court of Justice was faced with a particularly difficult issue, as national proceedings before specialized judges in fiscal fraud affairs, aimed, as such, to protect also EU financial interests, were suspected to be not fully in line with the right to effective judicial protection enshrined in Article 47 CFREU.

Looking at the passage of RS reported above, one can note that the reconstruction follows a peculiar order[78]. Primacy as a conflict rule is in fact the final point made by the Luxembourg Court, while the starting argument is construed on the duties of Member States to cooperate to ensure full effectiveness to EU law. Primacy is then restated also in the face of contrasting rules even of a constitutional nature and then connected with autonomy of EU law. Finally, with a restatement of the duty to ensure full effect of EU law, is then specified that, when the European provision at issue has direct effects, then national judges shall set aside, without waiting any internal procedure even of a constitutional nature, the internal contrasting rule.

The passage in which the Court of Justice seems to rely especially on the principle of primacy as such – without calling into question a specific provision – is the one dedicated to the discard of internal rules even of a constitutional nature[79]. Primacy was born as a principle, not embedded in an explicit provision of the Community or Union Treaties, therefore such reference seems coessential to its very nature. Nonetheless, the passages where the Court of Justice seems more in need to invoke primacy as such are those where the need to keep national common judges immune from any negative influence on their freedom to act as Union judges, i.e. to ensure that EU norms with direct effect prevail over national norms, with the freedom to refer, where deemed opportune, preliminary questions to the Court of Justice. Where the obstacles come from national constitutional rules, namely from rulings of constitutional courts, the CJEU has resorted to primacy in its essence.

Two remarks must be made, that help to understand this point. First, one must keep in mind that, as already pointed out, primacy as such was called into question by the referring judges in Euro Box (the reference in RS seems incidental and the question of interpretation are referred to identified provisions). Second, in Euro Box, which is, of the two considered, the one published slightly before, there is a point of the operative part where the Court of Justice offers its interpretation directly of the principle of primacy as such, precisely with reference to the problem that common judges cannot be precluded by a decision of their constitutional court of disapplying the orientation of such court, if they are in contrast with those of the Court of Justice.

The citations analysis of Euro Box and RS has been useful to highlight and confirm the following essential points:

• The story of primacy is reconstructed by the Court of Justice, especially in delicate judgments where difficult institutional issues are at stake; in this cases, traditional foundations of primacy are connected with relatively recent acquisitions of the case law, like those in the field of autonomy of the EU legal order;
• The Court of Justice has adjusted its approach on primacy over time, reinforcing in the last period some elements that sound rather intrusive in national legal orders (take the setting aside of constitutional courts’ rulings in contrast with orientations of the Court itself), namely in respect of the rule of law backsliding;
• The language used for expressing primacy of EU law has also evolved over time and it has become more direct, namely as regards, again, the problem of constitutional courts’ decisions in contrast with EU law and especially with the case law of the Court of Justice: in such situations the Court is inclined to resort to primacy as such.

The case-law dealing with general or specific aspects of primacy is very vast and citations analysis may be of help in many other situations. Without entering in a whole analysis of citations and just by way of example, one may also think to Melki and Abdeli. According to a widespread opinion, in that judgment the Court of Justice softened (if not overruled) the Simmenthal jurisprudence, as regards the obligation of common judges to set aside national norms in contrast with EU law provisions without waiting for a national procedure to formally annul or declare the former unconstitutional, by stating that

‘In so far as national law lays down an obligation to initiate an interlocutory procedure for the review of constitutionality, which would prevent the national court from immediately disapplying a national legislative provision which it considers to be contrary to EU law, the functioning of the system established by Article 267 TFEU nevertheless requires that that court be free, first, to adopt any measure necessary to ensure the provisional judicial protection of the rights conferred under the European Union’s legal order and, second, to disapply, at the end of such an interlocutory procedure, that national legislative provision if that court holds it to be contrary to EU law’.[80]

The passage reported, which is the one where the most innovative part of the judgment is formulated, does not contain citations. By contrast, Simmenthal is cited before, where its content is formally restated, while just after the reported passage other cases, regarding the exclusive power of the Court of Justice to declare an EU law act void, is cited. [81] That an innovative paragraph does not contain citations cannot be considered problematic. What is, on the contrary, not easy to explain is the fact that the Simmenthal rule has been formally restated and that no argument has been submitted to justify the softening (or re-reading, or overruling) of that jurisprudence. In doing so, the citation of Simmenthal has not been avoided, but it has been kept at distance from the formulation of (what many intend as) the new rule.

5.   Concluding remarks

Limited attention has been devoted to citations by the legal doctrine, thus far[82]. Nonetheless it has been noted, for instance, that in Francovich,[83], when dealing with State liability for violation of EU law, the solemn citation of the ‘fantastic four’ – van Gend en Loos[84], Costa v ENEL[85], Simmenthal[86] and Factortame[87] – marked the tie of the judgment with those illustrious precedents, and thus, Francovich was posed in the long vein of primacy. It was also stressed that the ‘collective appearance’ of the four landmark rulings in question is ‘otherwise very rare’.[88] With the dock of the Court of Justice increasingly overcrowded and institutional tensions involving the basics of EU law – starting with primacy – raising to pitches that were unexpected a decade ago or so, there is no need to rely on verified statistics to understand that today the citation of great classics has become recurrent (even if, probably, the ‘fantastic four’ collective appearance is possibly still confined to cases where the invocation of all of them appears justified).

When studying EU law and, inevitably, large or small numbers of CJEU’s judgments, one cannot escape from noting that the citation of precedents is widely used. Making a so complex little universe of case law grow and keep its coherence, to the standard deemed desired and achievable, is not an easy task and one should enter in touch with the way they do it in Luxembourg.

Examination of citations should be thus considered as a part of the textual analysis of the decisions of the Court of Justice and the foregoing presentation demonstrates it can be useful to highlight aspects that might sometimes rest hidden, or to confirm and corroborate hypotheses and results obtained thanks to other techniques (related to other aspects of the text of the judgments, or emerging from interpretative reconstructions, or from other documents or sources). Its use presents advantages, but also some risks that must be borne in mind. To conclude the present pages, I indicate two attitudes that can be combined for a fruitful use of the technique presented here, each of which, however, should be used carefully, as they may otherwise lead to some exaggerated or paradoxical results.

First, originalism is an attitude that can be fuelled by citations analysis, in particular, by an excessive reliance on it. If originalism as an interpretation approach to normative texts stresses the need to be faithful to their original meaning, a kind of similar attitude could exist towards the case law of the Court of Justice. Under this perspective, the interpreter may tend to privilege a specific meaning of a precedent considered ‘real’ and ‘original’, with reference, for instance, to the context in which the ruling in question has been adopted. In other words, if the Court cites that ruling, it must be because it wants to use that specific and original meaning. This attitude runs counter the impression – recurring on many occasions – that the Court of Justice often refers to a present interpretation of precedents, especially if they are old ones. It has already been said, for instance, that big classics are used with reference to the specific context of the rule of law backsliding, meaning, of course, that their proper significance cannot be turned upside down, however a too strict originalism could prevent the interpreter to catch the actual idea the Luxembourg Judges want to carve out from that ruling. Moreover, when citations are used with some erraticism, with the reference not being to the real case where the CJEU made the point of law in question, originalism can be disguising again (or should be, at least, referred to the real original precedent to which the point of law in question has to be referred back).

Second, another idea strictly related to the recourse to citations analysis has to do with systematic interpretation. It has been stated, with reference to EU normative sources, that ‘[s]ystematic interpretation is based on the premise that the legislator is a rational actor. This means that the authors of the Treaties are assumed to have established a legal order that is consistent and complete’ (emphasis original).[89] Paraphrasing this statement, it could be said that systematic interpretation of the case law is based on the premise that the Court of Justice is a ‘rational actor’ and that is assumed that it produces a jurisprudence that is consistent and tends to be complete. This assumption normally lies at the roots of the approach of commentators, who work to understand the judgments, discover possible inconsistencies inside the legal reasoning of each one of them and, more broadly, within the corpus of the case law. Moreover, they tend to prefer understandings that ensure that each ruling and each line of case law plays a clear role in the context of the whole jurisprudence and that some basic principles are identified, inspiring the reconstruction of the whole system. Carrying out citations analysis can serve the objective of this approach. However, it could, at the same time, exacerbate little problems, that may be due to light inconsistencies, and/or are related to not particularly significant drafting inaccuracies.

To counterbalance the risks of both originalism and overestimation of apparently anti-systematic elements, citations analysis should not rest a technique significant per se, but the results it offers are always to be read under the light of the developments of the EU legal order and of the case-law of the Court of Justice. In other words, changes in the understanding of precedents, or of entire lines of case law, can be signs of vitality of the legal order and of the evolution of its features and of its historical mission. The art of the interpreter should include the ability to distinguish basically three different cases in this respect. The first one regards inaccurate citations, when their impact can be light, thus their systematic significance is low. Careful consideration should be given to the second case, where inconsistencies or unjustified ruptures occur, and the Court of Justice should be called upon take its responsibility, in particular to better state reasons for the solutions adopted. In the third case, incoherencies in citations may disclose the unfolding of possible developments of the case law, that deserve attentive consideration.

In conclusion, the aspiration for a more case-specific argumentation, that would be more sober and better elaborated at the same time[90], collides with systematic choices of the Court of Justice that are in place at least since the beginning of the eighties, that tend to privilege the idea that the jurisprudence of the Court is speaking altogether each time, recalling consolidated arguments, that could not lead but to the result finally adopted by the Court itself[91]. Nowadays it cannot be denied that the heavy use of literal repetitions (accompanied in the majority of cases by the citation of at least the last judgment containing the same expression) helps also in speeding up the drafting exercise and makes research engines and machine learning work easily. The turnover of this trend seems improbable, thus EU law scholars should become more and more attentive to citations and to the hidden logical processes conducted – consciously or not – underneath their surface …still without abandoning the somewhat romantic desire of a different way to state reasons in judgments.