Has the Court of Justice embraced the language of constitutional identity?

In February 2022 the Court of Justice delivered its landmark judgments on the validity of the Conditionality Regulation. In its twin decisions, which dismissed Hungary and Poland’s actions for annulment, the Court confirmed that that the legal basis of the Regulation was adequate, that the new instrument does not conflict with Article 7 TEU, and that it adequately guarantees legal certainty. In order to reach those conclusions, the Luxembourg Court used also the language of ‘identity’. It did so in two parts of the ruling, first when it reflected on whether a horizontal conditionality mechanism based on the rule of law could be lawfully adopted on the basis of Article 322 TFEU; and then when it responded to the Hungarian and Polish claims that the rule of law and the values of Article 2 TEU lack legal relevance and have only a political and symbolic impact. In both points the Court used the same formulation: Article 2 TEU, according to the Court, contains ‘values which (…) are an integral part of the very identity of the European Union as a common legal order’ (paras 127 and 232 of the Hungarian case). The values of Article 2, we could say, are the ‘constitutional identity’ of the Union, and as the Court also holds in the judgments, the Union ‘must be able to defend those values’, to defend its identity, though always within the limits of the powers assigned to the institutions by the Treaties.
There is then a second ‘identity’ dimension in the Court’s decisions, which links the discourse on the EU’s identity to the national identities of the Member States. The Court refers in para 233 of the Hungarian case to Article 4(2) TEU, which asks to the EU to respect the national identities of the Member States, and argues that it derives from that provision that the Member States ‘enjoy a certain degree of discretion in implementing the principles of the rule of law’. Yet, the Court immediately adds, it does not follow from Article 4(2) that ‘the obligation as to the result to be achieved’ (i.e., the respect for the rule of law) varies from one Member State to the other. On the contrary, while the Member States maintain separate national identities, they ‘adhere to a concept of “the rule of law” which they share, as a value common to their own constitutional traditions, and which they have undertaken to respect at all times’.
The first analyses of the decisions (Pohjankoski; Kirst; Sachetti; and, most notably, Faraguna and Drinóczi and Bartole and Faraguna) have already discussed the Court’s use of the identity language and have highlighted in particular two aspects. First, they noted the novelty of using the concept of ‘identity’ with reference to the EU legal order, and the – admittedly modest – elaboration of a ‘EU constitutional identity’. Second, they underlined the rare clarity of the Court in setting out what counts – with a little wordplay – as a ‘constitutional’ ‘constitutional identity’, that is to say, what national identity claims are acceptable at the EU level, namely claims that are not in contrast with the common EU values of Article 2 TEU.
Not as much attention (but see Selejan-Guțan and Hogic) has been dedicated to two other recent developments before the Luxembourg Court: the decision of the Court in the RS case, related to the independence of the judiciary in Romania and the conflict between the Court of Justice and the Romanian Constitutional Court; and an Opinion of Advocate General Emiliou in a case referred by the Latvian Constitutional Court. Both cases also deal with questions related to constitutional identity, though exclusively from the perspective of national constitutional identity. What is remarkable in both the judgment and the Opinion is the level of detail with which the Court of Justice and the AG deal with the national identity questions, something that stands in contrast with the usual reluctance of the Court in elaborating on Article 4(2) TEU, as also highlighted by AG Emiliou in his Opinion (see also Faraguna).
In the Romanian case, the Court develops the point on national identity on its own motion, taking into consideration the broader background of the case pending before the referring court. It assesses in particular the previous refusal of the Romanian Constitutional Court to give effect to another preliminary ruling decision of the CJEU, a refusal that was grounded also on constitutional identity arguments. The Court develops an unusually clear analysis of how Article 4(2) TEU may (and may not) be used. In para 69, it holds that the Court of Justice may under Article 4(2) be called to determine whether a EU law obligation undermines the national identity of the Member States. At the same time, the Court strongly states that the national identity provision ‘has neither the object nor the effect of authorising a constitutional court of a Member State … to disapply a rule of EU law, on the ground that that rule undermines the national identity of the Member State concerned as defined by the national constitutional court’. While the statement as such is not surprising, in the sense that the established case law of the Court had never endorsed such a possibility, the firm stance of the Court is, to my knowledge, unprecedented. The Court in any event does not stop there. It continues by stating that if a national court considers that a secondary EU law provision infringes that national identity of its Member State, that court should stay the proceedings and refer a preliminary question of validity to the Court of Justice. The latter will then assess whether that secondary EU law norm should be annulled because it infringes the obligation to respect the national identity of the Member States. Again, this might not be a groundbreaking conclusion (among others, Di Federico had already convincingly explained that the national identity clause serves as a benchmark for the validity of secondary EU law), but the Court reiterates it with uncommon clarity.
Few days after the ruling of the Court of Justice in RS, AG Emiliou published his opinion in case Cilevičs and others. This is a preliminary reference from the Latvian Constitutional Court dealing with a completely different issue: whether national legislation requiring higher education institutions to offer courses only in the national language is compatible with EU law. One of the possible justifications presented by the Latvian government for this restriction to free movement was that the national authorities wished to protect and promote the national language, which is in turn considered to be part of the Latvian national identity. In order to tackle this argument, the Opinion offers a broad discussion of the roles and functions of Article 4(2) in the EU constitutional framework. AG Emiliou starts by acknowledging that until today, ‘the Court has not elaborated on the concept of ‘national identity’ or on the nature and scope of the ‘national identity clause’ set out in Article 4(2) TEU’. In particular, it remains unclear whether and to what extent ‘Article 4(2) TEU may be interpreted as introducing a horizontal or general clause that Member States may invoke in order validly to claim derogations from the EU rules’.
AG Emiliou puts forward five points on the issue. First, he argues that Article 4(2) TEU has ‘a dual nature’. On the one hand, it can work – in line with what seen earlier in RS – as a benchmark or parameter for the validity of EU law and acts. On the other hand, Article 4(2) TEU also requires to EU institutions, including the Court, to take ‘into account Member States’ national identities when interpreting and applying EU law’. As a second point, the Opinion deals with the scope of Article 4(2). The AG prefers a narrow reading of the provision, arguing that it can only be invoked with regard to ‘core constitutional elements of a Member State’ (for a similar view, De Witte). While it is not for the EU or the Court to determine what elements form part of that core of national identity, the Member States’ discretion is limited; otherwise, Article 4(2) would become ‘an all-too-easy escape clause from the rules and principles of the EU Treaties’. A third element discussed in the Opinion is in line with the considerations offered in the decisions on the Conditionality Regulation: in clear terms, the AG states that the national identity claims raised by a Member State must always be compatible and in line with the values affirmed in Article 2 TEU. Considering that these values are ‘common’ to the Member States, ‘Article 4(2) TEU cannot be considered to derogate from [Article] 2 … TEU’. In his fourth and fifth considerations, the AG then sets out how, in his view, a national identity claim should be resolved when the question concerns the compatibility of a national measure with EU law. AG Emiliou argues that a proportionality test should take place and that the Court of Justice should, in cases such as those referred by the Latvian court, leave the final decision to the competent national court, rather than reaching a ‘firm conclusion’ on proportionality.
All considered, the Opinion of the AG is highly stimulating. It relies on the Court case law (in particular Runevič-Vardyn and Wardyn, Bogendorff von Wolffersdorff, and Sayn-Wittgenstein) and existing scholarship (among others, Di Federico, Millet, Schill and von Bogdandy) in the attempt to formulate a more precise doctrine to operationalize and manage national identity arguments. In some senses, the Opinion seems favourable to possible national claims, for example when it argues that the proportionality assessment should be left to the national court. At the same time, the AG clearly reaffirms the pre-eminence of the common values provision of Article 2 TEU on the national identity clause of Article 4(2) TEU, and that the latter cannot be used as a unilateral ‘escape clause’ from EU law.
Now, the three judgments and the AG Opinion should probably not be considered a total paradigm shift in the Court’s approach to constitutional identity. On the EU side, in the decisions on the Conditionality Regulation, the references to identity were not truly decisive for the Court’s conclusions, but mostly served to highlight the constitutional dimension and impact of the judgment; and whether the concept of ‘EU constitutional identity’ is destined to play a larger role in the Court’s case law and what precise consequences should be attached to it is entirely to be determined. On the Member State side, in both RS and Cilevičs and others the Court and the AG are to a large extent repeating a series of principles that have been taken for granted for a long time.
But even without direct cross-references, the Court’s judgments and the AG Opinion seem to talk to each other and to suggest a more direct and explicit approach of the Court of Justice to national and constitutional identity claims. Until today, and again as acknowledged by AG Emiliou, the Court had been reluctant to develop a clear approach to identity claims based on Article 4(2) TEU. In some cases, it plainly refused to engage with identity questions, as in the Taricco II decision. In others, even when accepting claims based on national identity and national diversity, it refrained from making Article 4(2) the focal point of its decisions (as for example in Bogendorff von Wolffersdorff). As I have argued in an earlier contribution (and see also Spieker and Faraguna), the reluctance was probably motivated by the fear that a more precise, and possibly narrow, definition of the identity clause could have exacerbated, rather than resolved, conflicts with national constitutional courts.
The Court seems now on a different trajectory, one in which it is mostly explicitly embracing the language and concept of ‘identity’, and seems ready to develop a clearer doctrine and approach to Article 4(2) TEU. In all the cases discussed above, there is an attempt to clarify the scope, meaning and impact of Article 4(2) and provide ways for national courts to operationalize it within the boundaries of the EU constitutional framework. This renewed approach could be prompted by two connected considerations. First, the Court may have come to the realisation that in order to fight ‘abuses’ of constitutional identity (as described by Fabbrini and  Sajó, and Pech and Kelemen), silence and lack of engagement do not work as an answer. It is actually important to show what the correct and permissible use of identity is, as that helps, in turn, in defining an otherwise loose and vague concept of abuse. Second, the Court seems to be aware of the need to maintain a fair balance between commonality and diversity. The strengthening of commonality with the definition of the Union’s own identity in the Conditionality Regulation decision is balanced by an opening – at least to a degree – to national identities, in the sense that the Court in RS and the AG in Cilevičs and others explicitly accept the possibility that the national identity clause can be used as a benchmark for the validity of EU law and as a ground to derogate from EU law. The formulation of a more precise doctrine of ‘constitutional’ national identity may incentivize national courts to actually rely more often on Article 4(2), though the Court of Justice stresses the importance to proceduralize those claims: it is essential that the identity claims are raised in the context of the preliminary reference procedure and that national courts exercise sincere cooperation.
Only time will tell whether these first indications of the Court will be consolidated in a robust and structured ‘identity doctrine’. So far, the developments are promising, and I believe that the Court has more to gain than to lose from a more thorough engagement with national identity claims. Captured constitutional courts, like the Polish one, will not be convinced by the Court of Justice’s approach, but again the formulation of clearer criteria makes it easier to define what should be considered an abuse of national identity (see Scholtes and again Bartole and Faraguna). Other (constitutional) courts that have proved to be keen to use identity language – the Bundesverfassungsgericht immediately comes to mind of course, but also the Romanian Constitutional Court has recently used the concept in a rather confrontational fashion, as noted above – are invited to define constitutional identity in a dialogical process with the Court of Justice (see on this the suggestions of Martinico). The invitation to constitutional dialogue via the preliminary reference may also contribute to definition of the Union’s own constitutional identity, a concept which the Court has put forward in the Conditionality Regulation decision. After all, those values derive from and are shared with the Member States, and their definition is a common enterprise in the EU composite system.