National Courts and the construction of Europe: United in diversity

The decisive role played by the CJEU of the European Union (CJEU) in European integration is often, and rightly, highlighted ¹ .

As Walter Hallstein, the first President of the Commission of the European Economic Community, wrote, the Community, and henceforth the Union, has “no direct power of coercion, no army, no police. Its only instrument, its only weapon, is the law that it lays down”; to which he added: “its mission would be threatened to the utmost and, ultimately, defeated, if this sole means of implementing the Community’s objectives lost its binding and uniform character in all the Member States.” ²

In these circumstances, the CJEU could legitimately see itself as being invested with a particular mission, which very early on led it to affirm, absent any explicit basis in the treaties, the principles of direct effect ³ and primacy ⁴ . The two principles reflect the requirement inherent in the European project of ensuring that the Union isn’t bogged down in a system with excessively random or contingent applications. It also defined the contours of this specific, autonomous and integrated legal order that is the European Union and which, in its relations with those of the Member States, responds to an original logic, different from that which traditionally dominates the relations between intermunicipal law and municipal law ⁵ . From the 2000s onwards, the CJEU has continued its creative work by developing what might be called a Europe of rights and values, with the extension of the Union’s competences, the widening of the scope of its law, the emergence of European citizenship and the enshrinement of the Charter of Fundamental Rights – the implications of which now extend well beyond the economic sphere initially targeted by Europe’s founding fathers.  

The Court has thus been one of the main drivers of European integration, largely due to a form of activism it has constantly been reproached for.

This should not, however, detract us from the equally decisive role played by national courts. As the guardian of the treaties, the CJEU certainly sets the proper interpretation of EU law, but its action would be null and void if national courts were not there to ensure its effectiveness. In fact, the judges are responsible, by virtue of a sort of “jurisdictional subsidiarity” ⁶ , of applying it to the concrete disputes before them and of bringing to life, in the field, the logic of pre-eminence and integration which are at the heart of European construction. It should also be noted that the vast majority of decisions rendered by the Court are at the initiative of national courts, insofar as they decide to make use of the preliminary ruling procedure provided for in Article 267 of the Treaty on the Functioning of the European Union (TFEU).

However, the courts of the Member States are not simply executors, subordinated in a hierarchical and vertical logic to rulings coming from above. On the one hand, as ordinary judges of Union law, they are also those who are first and foremost faced with the difficulties that the application of this law can raise in practice. On the other hand, insofar as they are also the guardians of the national constitutional orders, from which they derive their authority and legitimacy, national courts must prevent the “conflicts of primacy” which are a constant threat, by elaborating mechanisms capable of ensuring a harmonious coordination between the legal orders of the Member States and of the Union. In either case, they can resolve these difficulties either spontaneously or by referring them to the Court, thereby giving it the means, where they consider it appropriate, to enrich or modify its case law. This is an essential function, because the tension between unity and diversity, which is the essence of European integration, can only be overcome through a real dialogue and not the establishment of unilateral solutions, which are often irrelevant and therefore doomed to failure.

The tensions which have recently marked the relations between the CJEU and certain national courts, several of which, such as the Polish Constitutional Tribunal, have not hesitated to directly question the primacy of Union law, are not the occasion to question the essence of the European project, but to reflect on ways to strengthen this dialogue and the mutual understanding on which it must be based, and to reaffirm the essentially collective nature of the work to which the Court and the national courts contribute together. 

I propose to briefly outline the way in which national courts, in particular the French Conseil d’État, have gradually established themselves as central players in the European integration (1), before mentioning a number of avenues for reflection that I believe should be explored in order to further develop the dialogue between them and the CJEU of the European Union (2).

I. National courts guarantee the effectiveness of Union law while ensuring that it is harmoniously linked to their constitutional rules

A. National courts have gradually abandoned their initial reluctance to allow Community law to penetrate their domestic legal spheres and have fully assumed their role as ordinary judges of Union law

The hesitations of national courts were undoubtedly caused by the revolutionary nature of the principle of primacy, which upset the conception that many judges had of their institutional role under the separation of powers, while also clashing with their attachment to a pyramid of norms at the top of which sits the national constitution ⁷ . At the dawn of European integration, a minimalist conception of the role of the ordinary judge prevailed in most of the Member States: he or she was only the organ of the law and of the law alone, whose conformity with the Constitution or any other higher norm he or she could not control. The “outside” dimension remained the exclusive preserve of the legislative and executive powers. Under these conditions, it is understandable that the Belgian Final Court of Appeal or the Conseil d’État should take refuge behind the theory of the “loi-écran” to refuse to sanction the primacy and direct effect of EU law ⁸ . In dualist systems, this same minimalism led the Italian Constitutional Court to rule that a violation of EU law did indeed result in the State being in breach and liable at the international level, but “did not deprive the municipal law stating the contrary of its full effect” ⁹ . 

However, the full primacy of European law over municipal laws has gradually been recognized by all national courts, due to a profound transformation of their office ¹⁰ .

In France, the first stage was quickly taken by the Cour de cassation (i.e., the supreme court of the civil order) ¹¹ , only a few months after the Constitutional Council declared itself to be incompetent to review the compliance of municipal laws with France’s international commitments ¹² . The Conseil d’État took longer to lift its reservations, but finally decided to do so in view of the importance acquired by European law in the domestic order, the requirement for legal consistency and the concordant case law on this point of the various European constitutional courts and supreme jurisdictions. In the breach opened by its Compagnie Alitalia ¹³ decision, it thus recognized, one after the other, the complete primacy of the treaties ¹⁴ , regulations ¹⁵ and EU directives ¹⁶ over all internal legislative and regulatory provisions. 

The second stage consisted of drawing the consequences from these first decisions in order to deepen the integration of European law in the internal order. For example, when a municipal law is incompatible with European law, the Conseil d’État required the administration not to adopt any implementing measures ¹⁷ . It also required the administration to cease applying, on expiry of the transposition period, both written rules and unwritten principles of municipal law that were incompatible with the objectives of a directive that had not been transposed ¹⁸ . It also required the government to make use of the “de-legalization” procedure of Article 37 of the Constitution whenever a legislative provision encroaching on the regulatory domain disregards it ¹⁹ . The Conseil d’État also recognized the primacy of the general principles of European Union law ²⁰ and gave full effect to decisions of the CJEU given as preliminary rulings, even when they are based on questions referred by the courts of other Member States ²¹ . 

Finally, in a third stage, the Conseil d’État brought administrative case law in line with that of the CJEU. A series of decisions were issued to fill in the remaining gaps. With regard to non-transposed directives, which had in the past given rise to frictions with the Court, the Conseil d’État gave them full normative power by agreeing to review an individual administrative act in light of their unconditional and precise provisions ²² . In the area of liability, the Court’s case law led the Conseil d’État to abandon old solutions that were firmly rooted in our conception of the separation of powers, by recognizing the possibility of finding the State liable where both the law ²³ and a final court decision ²⁴ have disregarded EU law. 

All the courts of the Member States have, albeit at different speeds, followed a similar path, asserting themselves as the primary guarantors of the effectiveness of European law.

B. National courts have also developed mechanisms to prevent the risk of collision between the European order and their domestic orders

The European integration has in fact given rise to a precarious constitutional pluralism, in which the supreme norms of the national and European legal orders are in constant danger of colliding. For, on the one hand, the European Union legitimately aspires, as a supranational entity, to the primacy of all its normative production over the norms of the Member States, including constitutional norms, which in theory requires the national judge “to ensure the full effect of Union law by leaving unapplied, if necessary, on its own authority, any [national] provision that is contrary to it” ²⁵ . On the other hand, the national jurisdictions all logically consider that their constitutions take precedence in the internal legal order: as Ronny Abraham underlined, “[t]his supremacy is thus, as long as international society is based on the political fact of State sovereignty, a primary and unconditional truth” ²⁶ . This situation leads to a “normative aporia” ²⁷ since no solution is provided for determining who, in the event of a conflict, should have the last word. In these circumstances, national courts have a particular responsibility, alongside the CJEU, to prevent European pluralism from breaking down into a “complete cacophony” ²⁸ . 

To this end, several European courts have developed a “theory of equivalence of protection” that is likely to prevent most conflicts between supreme norms, which are particularly likely to arise in situations where a domestic act – a law or regulation – transposes a European norm, such as a directive, without any margin of discretion. The aim is then to prevent this internal act, whose substance is one with that of the standard it transposes, from coming into conflict with a guarantee enshrined in the national constitution. According to the theory of equivalence of protection, which originates in the Solange I decision of the German Federal Constitutional Court, the national court must first of all find out whether there is no equivalent guarantee in the European order – which is very frequently the case in the field of fundamental rights and freedoms. In such a case, the court will “translate” its constitutionality review into the European order and will first check the conformity of the domestic act in question with the European guarantee, if necessary by calling on the CJEU to intervene by means of a preliminary question. The European Court of Human Rights ²⁹ and the CJEU ³⁰ have themselves been inspired by this mechanism to create a reciprocal presumption of equivalent guarantees of fundamental rights between Union law and the European Convention for the Protection of Human Rights and Fundamental Freedoms. 

In France, the Conseil d’État adopted this approach in its Arcelor decision of 8 February 2007 ³¹ . But in its French Data Network decision, it also established the other side of this reasoning by allowing the defendant, in a dispute in which the conformity of a national standard falling within the scope of Union law with a European directive or regulation is challenged, to argue that a “rule of municipal law, even though it is contrary to the provision of European Union law invoked in the dispute, cannot be set aside without depriving a constitutional requirement of effective guarantees”. It is up to the administrative judge in this case “to investigate whether there is a rule or a general principle of European Union law which, having regard to its nature and scope, as interpreted in the current state of the case law of the Union judge, guarantees by its application the effectiveness of the constitutional requirement invoked”. If so, “it is for the Court, in the absence of any serious difficulty justifying a preliminary question to the CJEU, to set aside that argument before upholding the applicant’s plea, if appropriate”; Conversely, if “such a provision or such a general principle of the law of the Union does not exist or if the scope that it is recognized in the European legal order is not equivalent to that guaranteed by the Constitution, it is up to the administrative judge to examine whether, by setting aside the rule of municipal law on the grounds of its conflict with the law of the European Union, he would deprive of effective guarantees the constitutional requirement of which the defendant avails himself, and, if need be, to set aside the arguments of the petitioner before him” ³² .

In both cases, the courts try as far as possible to avoid any conflict between Union law and municipal law. But in circumstances where a collision is unavoidable, they reserve the right to give precedence to the constitution over European law. This is the meaning of the control of “constitutional identity”, which is often perceived as a threat brandished by the courts of the Member States towards the CJEU. The existence of such a counter-limit is justified in view of the nature of European pluralism. However, its activation is only legitimate on the condition that national courts fully play the game of cooperation and use it only as a last resort: this is the meaning of the efforts made by the Conseil d’État to bring French legislation on the retention of connection data in line with the European framework. However, this desire for coordination does not seem to be shared by all judges.

II. Cooperation between the CJEU and national courts would benefit from further development to ensure the relevance of Union law and further European integration

A. A dialogue between judges based on trust and mutual understanding is the first key to fruitful cooperation

Given the plurality of legal orders, sources of law and jurisdictions that characterize the European institutional system, the dialogue of judges has quickly become a necessity ³³ . This dialogue can be more or less formal and take place in discussion forums such as ACA-Europe, under whose aegis databases such as Dec.Nat and Jurifast have been set up, which respectively centralize the decisions handed down by national courts in Union law and the bulk of the preliminary questions referred to the Court. This dialogue also involves the close attention paid by national courts to the case law of Luxembourg and the other courts of the Member States, both when they are implementing Union law and when they are faced with problems which are certainly national, but could also arise in other countries. 

But its main channel remains the preliminary ruling procedure provided for in the Treaties with the aim of ensuring a harmonized interpretation of Union law. The use made of it by the national courts is a good indicator of the nature of their relationship with the Court. In this respect, we note that the national courts were very quick to make use of the preliminary ruling procedure, and the number of references continued to increase until at least the early 2000s. The example of the Conseil d’État speaks for itself in this respect, since it referred 18 questions between 1970 and 1999, 86 between 2000 and 2015, and between 10 and 13 in 2016, 2017, 2018 and 2019. Overall, these figures show a real willingness to coordinate on the part of national courts. 

However, it should be noted that the courts of the Member States, when they make use of the preliminary ruling procedure, nowadays adopt a partnership approach rather than a hierarchical one, in which their status as ordinary judges of Union law is fully affirmed. This can be seen, on the one hand, in the margin of appreciation that they are careful to retain as to whether or not to refer a question to Luxembourg. In its 1982 ³⁴ CILFIT case-law, the Court adopted a particularly restrictive position, requiring that all questions be referred to it for which an answer is not “so obvious as to leave no room for reasonable doubt”. This case-law was probably justified at a time when all the courts of the Member States had not yet fully assumed their responsibilities arising from European integration, but it is much less justified today. It is understandable that most national courts have refused to apply it strictly, as a recent note commissioned by the CJEU shows ³⁵ . It is less understandable, however, that the Grand Chamber of the CJEU did not draw any conclusions from this when it was given the opportunity, very recently, to reassess the CILFIT criteria ³⁶ . National courts are aware of the role that falls to them in the construction of Europe, they know the law of the Union and are often in the best position to resolve the difficulties that may arise from its application, so much so that, as my colleague Jean-Denis Combrexelle noted, it seems to me that today, “institutional balance and no doubt wisdom dictate that the role of the supreme courts should not be confined to that of interpreting the obvious.” ³⁷

This additional autonomy that should be left to the national court seems all the more appropriate given that the time taken to make a reference for a preliminary ruling corresponds less and less to the time taken for justice to be done on the ground, which is constantly accelerating. It generally takes between one and two years for the Court to consider a reference for a preliminary ruling, which poses an obvious problem of timing, often playing a determinant role in the national court’s hesitation to submit the question. In comparison, the success of the priority question of constitutionality is largely due to the three-month time limit within which the Constitutional Council is obliged to rule. Beyond the question of the autonomy of the national courts, could the Court not better distinguish between cases justifying an accelerated procedure, whether they raise particularly sensitive questions or do not warrant the most thorough examination, and those subject to the normal procedure? An urgent preliminary ruling procedure was certainly created in 2008 ³⁸ , but it clearly does not manage to respond effectively to the acceleration of judicial time. Could we not also imagine that the Court could intervene as amicus curiae before national courts when a question of interpretation of Union law arises? Some thought must be given to resolve this problem.

In any event, the partnership approach referred to above is also expressed in the “uninhibited” attitude of the national courts ³⁹ , which take on the clothes of a “zealous interlocutor” when they seek, by referring a case for a preliminary ruling, to maximize the scope of the principles of direct effect and primacy ⁴⁰ , as well as those of a “proactive interlocutor” who, by asking questions, seeks to guide or develop the Court’s case law. This attitude is a sign of the maturity of the national courts which, as ordinary judges of Union law, contribute not only to its application but also to its development. The Conseil d’État‘s sometimes rough dialogue with the Court on the subject of the retention of connection data is a perfect illustration of this: by giving the Court the means to clarify its Tele2 Sverige case law ⁴¹ in order to take account of the reality of intelligence and criminal investigations in France and many other countries, the Conseil d’État wished to play an active part in defining a protective, realistic and effective European legal framework. It has done so by resolutely rejecting the logic of opposition, or even rupture, to which the government was calling when it asked it to carry out an ultra vires control, favoring instead the path of dialogue, seen as the only constructive path. Its decision in French Data Network ⁴² thus feeds the Court’s reflection, but also, more broadly, that which is underway with a view to drafting the new regulation that will replace the 2002 “e-privacy” directive ⁴³ .

It should be stressed in this respect that the tensions between the Court and the national courts often stem from a problem of text before being a problem of case law. The e-Privacy Directive, drafted at a time when space connection data was yet to gain the attention it has today, is notoriously obsolete. The 2003 directive on working time ⁴⁴ does not sufficiently address the situation of military personnel in certain countries such as France. The temptation for Member States is then to ask national courts or the Court to repair these defective texts. But should it not be up to politicians to sit down and renegotiate these texts?  Otherwise, by shifting such responsibility to the judges, we sow the seeds of discord. 

B. A renewed use of the principle of subsidiarity and the concept of national margin of appreciation could in this respect ease the tensions recently observed between European judges

We cannot turn a blind eye to the tension – if not outright hostility to the Court and to the Union – shown by certain decisions handed down recently by national courts. This is illustrated by the active use of the ultra vires review by the Czech ⁴⁵ , Danish ⁴⁶ and German ⁴⁷ constitutional courts, and of course by the recent decision of the Polish Constitutional Tribunal ⁴⁸ , which has given rise to comments that are not only harmful but also legally incorrect. On this point, we must be clear: the fact that national constitutions take precedence in the internal orders of the Member States does not mean that they can disregard the fundamental principles of the rule of law through authoritarian excesses. When such principles are openly flouted, it is to the credit of the Union’s institutions, in particular the Court, to adopt the firmest attitude. To accuse them of stepping outside their mandate and of being excessively rigid in such cases is dangerous and attests to a clear bad faith. 

These decisions remind us, however, that national courts are ultimately free to decide whether or not to apply the Union law, and that, just as they have been the key architects of European integration up to now, they could be the workers in its deconstruction tomorrow. A case like Poland’s certainly calls for a political response, to which the Court can add little. But in many other cases, faced with the growing influence of Union law, it is likely that national courts will be all the more cooperative if they are allowed a certain amount of flexibility in its interpretation and application. It is true that in some areas, particularly economic, the uniform application of Union law is essential. In other areas, in particular those under the third integration pillar, the Union’s common framework can accommodate certain divergences without calling into question the requirement of uniformity. 

This could be achieved, on the one hand, by recognizing that the Court does not have a monopoly over the interpretation of EU law. Such recognition would mean, first, accepting the reality of the division of roles between the Court and the national courts. Second, it would give full effect to the spirit of the second paragraph of Article 4 of the Treaty on the European Union, which requires the Court to respect “the national identities [of the Member States], inherent in their fundamental political and constitutional structures”.  Finally, it would sanction the particular position of national courts, which are required to show deference to the Union but also, for obvious reasons of democratic legitimacy, to the choices made by their own legislators. Responding to constitutional pluralism, the resulting “interpretative pluralism” would be ordered around a shared ethic, whereby judges would only allow themselves to depart from the Court’s case law for sound and duly reasoned considerations ⁴⁹ . 

Symmetrically, the Court could also draw more inspiration from the case law of the European Court of Human Rights, which is obliged “to show restraint in the exercise of its control of conventionality, whenever this leads it to assess an arbitration carried out in accordance with democratic procedures” ⁵⁰ . In this spirit, it modulates the margin of appreciation it leaves to Member States according to the importance of the rights in question and the consensus that exists or not, at a given moment, as to the scope they should be given. It is true that the norms stemming from the European Convention for the Protection of Human Rights and Fundamental Freedoms, which is based on the interpretation by the courts of principles that are both very general and plastic, lend themselves more to this kind of exercise than EU law, particularly secondary legislation, which is becoming increasingly precise, detailed and descriptive, including in the ways in which Member States can use this or that option left open by the text. Such an approach could, however, be envisaged by the CJEU in third-pillar matters, as well as when certain balances between fundamental rights and freedoms are at stake, which are usually closely linked to the political, legal and cultural traditions of the Member States. 

I believe that these avenues for reflection must be explored so that, at the cost of a minor notch in the principle of uniformity of Union law, any disagreements between judges can be seen as an asset and not a threat, and so that courts can continue together, united in their diversity, their common work in the service of an ever-stronger Union.

1. Text written in collaboration with Guillaume Halard, administrative magistrate, chargé de mission with the vice-president of the Conseil d’État.
2. Speech to the European Parliament, June 1965 session. See, RTDeur, Dalloz, 1965, p. 250.
3. CJEU 5 February 1963, Van en den Loos, 26/62
4. CJEU 15 July 1964, Costa v. ENEL, 6/64
5. See, R. Dehousse, ‘L’Europe par le droit : plaidoyer pour une approche contextuelle’, Politique européenne, 2000/1, no. 1, p. 63.
6. See, D. Simon, ‘La subsidiarité juridictionnelle : notion-gadget ou concept opératoire ?’, Revue des affaires européennes, 1998, esp. P. 84-85.
7. See, J.-M. Sauvé, ‘Le renouvellement du droit administratif sous l’influence du droit européen’, Mélanges en l’honneur de Bernard Stirn, Dalloz, 2019, p. 511.
8. CE 1 March 1968, Semolina Manufacturers Union, Lebon 149
9. Italian Constitutional Court, 24 February 1964, Costa v. ENEL
10. See, D. Simon, ‘Les exigences de la primauté du droit communautaire : continuité ou métamorphoses ?’, Mélanges en l’honneur de Jean Boulouis, Dalloz, 1991, p. 481.
11. Cass. 24 May 1975, Société des cafés Jacques Vabre, 73-13.556
12. Decision No 74-54 DC of 15 January 1975
13. CE 3 February 1989, Compagnie Alitalia, n° 74052
14. CE 20 October 1989, Nicolo, n° 108243
15. CE 9 September 1990, Boisdet, n° 58567
16. CE 28 February 1992, S.A. Rothmans International France, n° 56776
17. CE 24 February 1999, Association de patients de la médecine d’orientation anthroposophique et autres, n° 195354
18. CE 6 February 1998, Tête, n° 138777 and 20 May 1998, Communauté de communes de Piémont-de-Barr, n° 188239
19. CE 3 December 1999, Association ornithologique et mammologique de Saône-et-Loire, n° 164789
20. CE 3 December 2001, Syndicat national de l’industrie pharmaceutique, n° 226514
21. CE 11 December 2006, Société De Groot en Slot Allium B.V. , n° 234560
22. CE 30 October 2009, Mrs Perreux, n° 298348
23. CE 8 February 2007, Gardedieu, n° 279522
24. CE 18 June 2008, Gestas, n° 295831
25. CJEU 9 March 1978, Simmenthal, C-106/77
26. See, R. Abraham, Droit international, droit communautaire et droit français, Hachette, 1989, p. 35.
27. C. Malverti & C. Beaufils, ‘L’instinct de conservation’, AJDA 2021. 1194
28. J. Baquero Cruz, ‘The Legacy of the Maastricht-Urteil and the Pluralist Movement’, European Law Journal, vol. 14, No. 4, 2008, p. 414, quoted by C. Malverti & C. Beaufils, ‘L’instinct de conservation’, art. cit.
29. ECHR 30 June 2005, Bosphorus, No 45036/98
30. CJEU 3 September 2008, Kadi v Council, C-402 and 415/05 P
31. CE 8 February 2007, Société Arcelor Atlantique et Lorraine, n° 287110
32. CE 21 April 2021, French Data Network, n° 393099
33. See, B. Genevois, ‘Dialogue des juges ou confrontation sous-jacente ?’, in La Concurrence des juges en Europe. Le dialogue des juges en question(s), Ed. Clément Juglar, 2018, p. 19.
34. CJEU 6 October 1982, Cilfit and others,  283/81
35. Research note on the application of the Cilfit case-law by national courts whose decisions are not subject to judicial review under municipal law (https://curia.europa.eu/jcms/upload/docs/application/pdf/2020-01/ndr-cilfit-fr.pdf).
36. CJEU 6 Oct. 2021, C-561/19
37. See, J.-D. Combrexelle, ‘Sur l’actualité du “dialogue des juges”’, AJDA 2018. 1929.
38. This procedure is currently provided for in Article 23 bis of the Statute of the Court and Rule 104 ter of the Rules of Procedure of the Court (RP).
39. See, G. Odinet & S. Roussel, ‘Renvoi préjudiciel : le dialogue des juges décomplexés’, AJDA 2017. 740.
40. This attitude can be seen, for example, in the aforementioned decisions of the Conseil d’État Arcelor and Jacob of 31 May 2016, No. 393881
41. CJEU 21 December 2016, Tele2 Sverige AB, C-203/15
42. CE 21 April 2021, French Data Network, cited supra.
43. Directive 2002/58/EC of the European Parliament and of the Council of 12 July 2002 concerning the processing of personal data and the protection of privacy in the electronic communications sector (Directive on privacy and electronic communications).
44. Directive 2003/88/EC of the Parliament and of the Council of 4 November 2003 concerning certain aspects of the organisation of working time
45. Ustavní soud, 31 Jan 2012, Pl. US 5/12, Slovak Pensions
46. Højesteret, 6 Dec. 2016, Case 15/2014, Dansk Industri [DI], acting for Ajos / Succession Rasmussen
47. BverfG, 5 May 2020, 2 BvR 859/15, Sovereign bond purchase programme PSPP
48. Decision K 3/21 of 7 October 2021
49. See for these considerations, C. Malverti & C. Beaufils, ‘L’instinct de conservation’, art. cit.in fine.
50. ECHR 23 July 1968, “Belgian Linguistic Case”, no. 1474/62; ECHR 1 ^erJuly 2014, SAS v. France, no. 43835/11, § 154.