Living with Five Supreme Courts
John Bell 
Cite as Bell, J., "Living with Five Supreme Courts", (2015) 21(1) European Journal of Current Legal Issues.
Alan Paterson's research is sceptical about the so-called dialogue between national courts and European courts. In relation to the Luxembourg Court, he writes that, although a few British judges go to the annual meeting with the Court, "there does not seem to be much by way of intellectual exchange."  The structured dialogue in terms of preliminary references and the responses given by the Court which are then acted upon by the national court or used in interpretation seem akin to "sequential monologues", rather than dialogues.  Paterson is more convinced that the Strasbourg Court engages in dialogue. It does seem to change its mind as a result of dissent from the national courts, and national courts do seem to debate what it says. But all the same, the major theme of his work is the overly deferential approach of the national courts to the Strasbourg Court. Paterson's discussion raises an important issue about the character of interactions between supreme courts. If institutions are not hierarchically integrated, how do they interact and how do they remain coordinated? This is significant, not least from the point of view of the litigant and his or her lawyer.
This paper takes France as a suitable example of the continental European problem of integrating supreme courts, both at national level and at supranational level. France, like many countries, exists with a kind of confederation of supreme courts at national level and then can be further related to Courts in Luxembourg and Strasbourg. France operates with a weak doctrine of precedent - it is better to get the right answer than to get an answer which is predictable from the previous decisions of a higher court. So, coordination is a matter of persuasion, not command. My argument essentially is that, as Paterson suggests, we need to look at the mechanisms which promote "intellectual exchange", rather than simple "sequential monologues". For the most part in France (and I would argue elsewhere also), these are informal, as Mitchell Lasser suggests. 
A. The French Problem
France has never had a single "Supreme Court".  When the supreme civil and criminal court (the Tribunal (now Cour) de cassation was created in 1790, it unified the case law of the various regional courts, but it was not allowed to interfere with the administration. The Conseil d'Etat was created in 1799 and it was supreme in fields of administrative and tax law and gradually became recognised as an independent court. So within France there are two independent systems of courts, one in private and criminal law and one in administrative law, each with regional courts hierarchically integrated into a supreme court. Conflicts about the competence of each system of courts are resolved by the Tribunal des conflits created in 1848 and involving equal numbers of judges from each supreme court, presided over by the Minister of Justice, who has the casting vote. But this body does not have direct jurisdiction to resolve divergence in interpretation of the law. For example, the Tribunal des conflits would not have jurisdiction if there was a difference between the principles of liability governing blood transfusions in private hospitals (decided by the Cour de cassation) and those in public hospitals (decided by the Conseil d'Etat).
In 1958, the picture became more complicated by the creation of the Conseil constitutionnel as the constitutional court. Its functions were principally to rule on the constitutionality of parliamentary legislation before it was enacted and to determine when the Executive had constitutional power to legislate by decree without reference to Parliament. (It is also the electoral court for parliamentary and presidential elections.) Until 2010, this court was not connected directly to the Cour de cassation and the Conseil d'Etat, except that they are constitutionally bound to apply its rulings on whether a particular piece of legislation is constitutional. Often the Conseil constitutionnel declares a legislative provision to be constitutional only if it is interpreted in a particular way (the so-called "reserves of interpretation") . Such reserves must then applied by the administrative and civil courts. Since 2010, there has been a procedure whereby the constitutionality of legislation which has been enacted can be raised as a legal issue before administrative and civil courts and the Conseil d'Etat or the Cour de cassation can refer the issue for a preliminary ruling by the Conseil constitutionnel, which is then applied by the relevant public law or private law courts.  But this does not make the Conseil constitutionnel into a court unifying the law throughout the country. It is confined to making rulings on the interpretation of the Constitution, and does not rule on other legal norms.
This distinctiveness of court institutions is then reinforced by the distinctiveness of judicial personnel and of legal education. Judges for the three supreme courts are recruited in different ways. Judges for the civil and criminal courts (the ordinary judiciary) are recruited by a special competitive process and are trained at the Ecole Nationale de la Magistrature in Bordeaux. They then have a distinctive career as judges, prosecutors or in other public service roles (such as civil servants within the Ministry of Justice) under the supervision of the Conseil supérieur de la Magistrature. Judges for the administrative courts are recruited by separate competitions and are trained either at the Ecole Nationale de l'Administration (in Paris and Strasbourg) or by the Conseil d'Etat for those entering from the public sector. Their career is governed under the supervision of the Vice-President of the Conseil dEtat . By contrast, the members of the Conseil constitutionnel are not required to have legal training and are appointed for a fixed term. Some have been judges of the civil or administrative courts, but they are typically a minority. This contrasts with the more homogeneous legal education of the German judiciary. As a result, one cannot assume either common understandings or socialisation between the members of the different judiciaries that could create an informal basis for harmony between the legal systems.
Since 1953 and especially since the French courts made their first preliminary references in 1965, the Court of Justice of the European Union (as it is now called) has been a supreme court within the French legal system, but. Whereas the civil courts recognised the superiority of EC law over national statutes in 1975, it took the administrative courts until 1989. The Conseil constitutionnel takes the view that EU law does not prevail over the Constitution, but that it does over other statutes. The legal basis for the superiority of EU law within French law was settled by a constitutional amendment in 1992 creating art. 88-1. The French will have a judge (currently from the Conseil d'Etat) and an Advocate General (currently from the Cour de cassation), as well as some national judges working as référendaires in the cabinets of various judges, especially the French members of the Court. These various members of the French judiciary with experience of Luxembourg will typically return to take up senior positions within their respective judicial hierarchies in France. An example would be Jean-Claude Bonichot, currently the French judge at the Court of Justice, who is a member of the Conseil d'Etat and was, at one point in his career, a référendaire in the cabinet of ECJ Judge Galmot, before returning to a promotion within the Conseil d'Etat.
Whereas, formally, the ruling given in the preliminary reference sent by a French court is the only decision of the CJEU which binds a French court, and that for the individual case only. Nevertheless, the limited number of decisions from the CJEU and the deliberation which goes behind them gives them a much wider status.
France was slower in ratifying the European Convention on Human Rights. This came in 1974 and the right of individual petition came in 1981. The legal basis for the superiority of Convention law over national legislation is art. 55 of the Constitution which gives priority to treaty provisions over national legislation. But, because this is a provision of the Constitution, it does not prevail over the Constitution. But even then, one must distinguish between the authority of the Convention as a treaty and the authority of the judicial decisions of the European Court of Human Rights. As Braconnier wrote, "the absence of a kind of federal structure at the level of the Council of Europe obliges us to state that the decisions of the European Court of Human Rights have no defined place, in internal law, within the hierarchy of norms, apart from their relative authority as res judicata."  The French will have a judge at the Court, and some others will have experience of Strasbourg which is often brought back when the judge returns to France. But of course, there is a tension. The role of the French judge in Strasbourg is to be an independent person with an understanding of the French legal culture, not a representative of his system or country.  Furthermore, the ethos of the Strasbourg Court is different from the French legal system in terms of procedure, judgment writing and legal preoccupations. Examples in the past include Lecourt who was judge in Strasbourg before becoming a member of the Conseil constitutionnel. Also, in more recent times, Jean-Paul Costa was a President of Section in the Conseil d'Etat before becoming a Judge, then President of the European Court of Human Rights.
Although the action before the European Court of Human Rights is against the state and does not generate res judicata in a strict sense that the French courts must follow, there is a general tendency for French courts to follow decisions of that Court. As Braconnier noted, the influence of the Strasbourg Court has moved beyond introducing some technical adjustments and has now reached the scale of causing major movements for change within French law. All the same, the Strasbourg Court is concerned only with setting minimum standards and so it cannot create some new form of "European administrative law". 
So, the problem for the French lawyer is that there are five fairly autonomous supreme courts which he or she has to integrate into advice to a client or in litigating a case. Now, it is obvious that this is not resolved formally. There is no single court which has the final say. Luxembourg may disagree with Strasbourg on fundamental rights, and the three French courts could each have different views again. Gradually, they may align their positions, but this is the result of various informal mechanisms at work. As Lasser has astutely pointed out in his study of the French private law judges, it is very important to examine the informal discussions that take place in court in order to understand the reality of judicial reasoning, rather than to be focused (as common lawyers typically are) on the published text of decisions.
B. Influences Towards Coordination
Although the five supreme courts exist in very separate lives, they are subject to a number of common influences, which ensure that there is communication about what one court is doing to others.
Lawyers are not divided in the same way as judges. French avocats belong to regional bars and, since 1990, they can operate as part of firms covering more than one bar area. There has, therefore, been a growth in larger firms. These firms do not confine themselves to particular courts. (The relatively small number of cabinets which have the monopoly of cases before the Cour de cassation and the Conseil d'Etat also operate in litigation in other courts.) These lawyers will, rather, represent clients and their interests, which may cut across both administrative and civil courts. They may also plead where a preliminary question is referred to the Conseil constitutionnel or to the Court of Justice in Luxembourg, and they can clearly plead in cases before the Strasbourg Court. A firm may have a contract dispute before the commercial courts, an employment dispute before the labour courts, a tax issue or a planning issue before the administrative courts. Norms from EU law or the Convention may be involved. Because of the nature of litigation, the courts are not masters of their agenda, but respond to arguments presented by lawyers. Although courts can indicate what they think are acceptable arguments, there will be a level of pressure on them to respond to the arguments that the lawyers raise. So, for example, Lasser draws attention to research undertaken on the Conseil d'Etat in 2005. In that year, in 20% of cases decided by that court, French litigants raised issues involving the Convention, but 90% of foreign litigants did so . Such pressure forces the French courts to engage with the case law of the Strasbourg Court. Now that avocats also plead cases before the Conseil constitutionnel in the preliminary reference procedure, they are able to encourage the courts to confront the case law of other courts.
2. The incompleteness of law
A second feature is the incomplete, if not underdeveloped, nature of many of the branches of law in question. There is no substantive code of general administrative law. There are written rules of administrative court procedure (code de la justice administrative, dating from 2000) and there are specific codes dealing with particular sectors of the law (code forestier, code des mines, etc). So, on many issues, French administrative law has to look to private employment law in order to develop rules. So, for example, in the absence of precise rules on unfair dismissal, the doctrine of proportionality was used to determine the legality of civil service dismissals. 
The French Constitution contains very vague statements of fundamental values in the Declaration of the Rights of Man of 1789 and the Preamble to the 1946 Constitution, but there is no comprehensive statement of rights, and certainly no obvious statement of the hierarchy between rights. So constitutional law of fundamental rights needs to be developed. 
Equally, the fundamental rights laid down in the European Convention have needed considerable elaboration by the Strasbourg Court, which often leaves a margin of appreciation to the national courts (among other national public authorities). Even the heavily detailed European Union law leaves many general principles for interpretation. The interpretative openness of the basic norms to be applied encourages the judges to look around for analogies.
It is very common for the administrative judges to look to private law when dealing with relatively similar problems. This is not a new phenomenon. Most famously, the commissaire du gouvernement Romieu in 1895 looked to the French, Belgian and Luxembourgeois interpretations of the French Civil Code in order to work out the principles that should govern public law liability for work injuries to an employee in an armaments factory.  He chose to follow the lines of development abroad rather than in France and give the employee compensation without proof of fault on the part of the employer. That was then adapted by the Cour de cassation in 1896 in the case of injuries to a private law employee arising from a boiler explosion on a ship.  It is very common for the commissaire du gouvernement (now called the "rapporteur public") or for the reporting judge to discuss the case law in other courts. In the past, the public lawyers would discuss private law and the private lawyers would discuss public law.  Nowadays, both will discuss the case law of the Strasbourg and Luxembourg courts, as well as that of the Conseil constitutionnel. Such references may not always reflect that these courts feel bound by what another court has decided. Far from it. Each court retains its primary obligation to do justice to the parties as it sees it. But it is considered wise to examine the case law of other courts to seek inspiration. Although there are distinctive cultures and (sometimes) distinctive values which lead to different results, there is also a sense that the different parts of French law need to work together.
3. Europe as enhancing judicial status
The arrival of European law gave an enhanced status to the national judge in France. Traditionally, the judge was supposed to be a functionary within the legal order and had a relatively low status. In particular, since 1790, the judge was not supposed to call into question the law enacted by Parliament.  Since the ECJ decisions in Costa v ENEL and Van Gen den Loos, the French courts have been encouraged to disapply incompatible national law in favour of EU law. That has put the French judges into a different position relative to Parliament. The same is true of the European Convention. Both European sets of norms have enhanced the position of the national judge in relation to the legislature. The result has been that subordination to Europe has enhanced the status of the national judge.  Lasser is right to say
When French judges interpret European law, they do not merely 'apply' European law; they realise it. Although the European courts can periodically set or adjust the basic terms and requirements of European law, national courts constantly wield the interpretative power to shape and develop these legal norms according to their own perspectives and for their own purposes; in doing so, they establish the meaning of European law within their own jurisdictions…In effect, not only does the national judiciary determine in large measure what European law actually is and requires in the domestic legal order, it can also exercise that power most effectively when it increases the pertinence and expansiveness of European law. 
Lasser cites the cases on the role of the commissaire du gouvernement or the advocate general in judicial deliberations. Here the Strasbourg court eventually pushed the French to alter long-established rules of procedure in the interests of the appearance of justice.  But there is also the way in which the French courts have been willing to give an expansive interpretation to Convention principles. Even before the right of individual petition was allowed in France, the Conseil d'Etat expanded its concept of "general principles of law" to include a "right to lead a normal family life" with the result that this acted as a considerable constraint on deportation of foreigners.  Convention rights are, in an important sense, domesticated by the use of concepts like "general principles of law".
C. Informal Mechanisms
My argument is that, in order to understand the way courts are kept aligned, it is necessary to examine a number of informal mechanisms.
Much of what Paterson and others discuss by way of "judicial dialogue" takes the form of national courts writing judgments and supranational courts responding to them, generating a further cycle of correspondence through judgments. His comments based on interviews relate to the question of whether the concerns of English judges or the UK Supreme Court are taken into account by the supranational courts. But we should be wary of this as a benchmark for influence. In the first place, national governments are usually litigants before the Strasbourg or Luxembourg courts. In the latter court, they may well intervene in cases involving other jurisdictions. The direct and focused arguments of a government are likely to have a significant bearing on the result in the case more than earlier statements of national judges in other cases, however pertinent. Those statements will form a background that finds its way into the preparations for the Advocate General and the Reporter Judge. The model of the dialogue of judges underestimates the importance of the multi-party character of the conversations that a supranational jurisdiction undertakes. It also undervalues the informal ways in which the view of one court are understood and reflected upon in another court.
1. Overlapping personnel
With the exception of the relations between the Cour de cassation and the Conseil d'Etat, there is an overlap of personnel between the various supreme courts we are discussing.  Thus all the French judges and advocates general of the Court of Justice of the European Union have come from the Conseil d'Etat or the Cour de cassation. Similarly the members of the European Court of Human Rights have previous experience as judges in one of the corps of national judges. In the case of the CJEU, there is also a cabinet of advisers who are often drawn from the national judiciary. This enables the Strasbourg or Luxembourg court to have knowledge of how national judges think. It also provides informal channels of communication where research on national law or legal thinking is required as background to the development of a judgment at supranational level. But this communication works in the other direction as well. If national courts, as part of their research, want to understand the position of EU law or of Convention law to a particular issue, this can be checked with the French members of the court staff. For example, since there may be administrative judges in the Luxembourg court as référendaires, members of the Conseil can contact them informally when preparing their reports on a case. This may also happen in relation to the Conseil constitutionnel which finds it impossible to make a reference to the CJEU in relation to pre-enactment scrutiny of legislation. But it can try to discover through informal channels what the likely position of the CJEU is going to be on a topic and anticipate that.
A significant element in the overlap of personnel lies in the role of support for the judges. I have already mentioned the référendaires in the CJEU. In the French case, these will often be judges on secondment. They bring to the preparation of European cases their knowledge and experience of national judging and national law. The cabinet of a judge or an Advocate General will have a number of people with varied experience and also a range of contacts in the national system that can be tapped for informal information. The Cour de cassation has a service de documentation, des études et du rapport with specific sub-units on European and constitutional law. The Conseil d'Etat has the Centre de recherches et de diffusion juridiques. Both of these will provide support information for decisions and can follow the case law of Strasbourg or Luxembourg. The secretariat of the Conseil constitutionnel has an important role because the judges may well not be lawyers and they are often required by law to decide in a very short period. It produces background information, which will include summaries of case law from other courts. Its members are often judges, many from the Conseil d'Etat. The physical proximity of those two courts enables a close contact between the body which undertook pre-legislative scrutiny and the body which exercises post-enactment scrutiny. The Registry of the Strasbourg Court consists of lawyers and, in the French case, some will be judges.  This body supports judges in decision-making by providing briefing information and writing judgments. It has the task of ensuring consistency across the many different Sections in which the Court sits. So, apart from the potential overlap of the senior members of the different courts, there is a larger overlap among the less senior judges who advise the supreme courts.
2. Meetings and conversations
Paterson is particularly sceptical of the importance of meetings between judges of different courts. Mak  also notes that there are marked differences in participation depending on the personality of the judges. It is clear that there are a lot of gatherings and opportunities for formal and informal interaction between judges of the different French supreme courts and the European supreme courts.  Each of the French courts is also involved in a number of other links, notably with francophone courts and with the United States. There are also a number of associations at European level, not least the Consultative Council of European Judges.  The collaboration in such organisations provides a level of common understanding between judges. Our interest, like that of Paterson, is in the significance of the interactions.
Influence involves the shaping of opinions and perceptions. That occurs in a variety of ways, not just by reading materials which are presented within the framework of a formal judicial decision. The study of networks by Claes and de Visser  explains how associations and internet connections provide fora for the exchange of ideas amongst judges. Such networks facilitate face-to face meetings, periods of stay in a country or even observation in foreign courts. In terms of the use made of these interactions, those authors distinguish between "practical" and "authority" incentives, and this distinction is also helpful in identifying the influence of meetings and conversations.
"Practical" incentives are concerned with the importance of internationalisation for the actors involved in the judicial business. Participating in networks is not so much a matter of choice but of necessity to make sure that judges are still able to get the job done properly in a changing environment. The open borders within Schengen have accentuated the cross-border element in much litigation or criminal investigation. This makes knowledge about other legal systems a prerequisite to being able to dispense justice in an individual case. These considerations take on an extra dimension within the European Union and the expansive reach of EU rules.  Certainly the practical necessity has become important, and it has led to the placing of a number of "liaison" judges in different countries. Mak and others place relationships with Strasbourg and Luxembourg within the same dynamic of practical necessity in the contemporary world.
By contrast, "'[a]uthority' incentives denote membership of judicial networks for the purposes of borrowing or enhancing the participating court's authority and legitimacy".  Participation in a network may enhance the status of a court in the sight of judges from other countries or it may confer an additional aura of authority in its internal dealing within the legal system - being associated with others may enhance the authority of the domestic institution.  Particularly in the European context, they see the network as working alongside more formal structures, such as the preliminary reference to enable national judges to gain a more European perspective on the interpretation of common norms.  Supra-national courts such as the European Court of Justice and the European Court of Human Rights foster such informal networking events in order to create a common ethos within which individual decisions will be decided. Data is often available on the associations and conferences of which judges are a part. That does permit network analysis of who is meeting with whom and how frequently. The elite interviews conducted by Paterson  and Mak  have tried to ascertain the perception of the participants about the significance of such networking and its influence on decision-making. They conclude that national judges seek to influence supranational judges and vice versa. There is not a one-way street. Supranational judges gain a sense of the importance of particular issues at a national level and national judges gain some sense of why supranational judges take a different view. For example, Mak notes that there were divergent views on the utility of international exchanges amongst the judges to whom she talked. Some thought them important and others did not see how they would change much of what judges ordinarily did.  Certain judges are more likely to take up these opportunities than others. But the conclusions both reach is that the number and variety of such exchanges has increased.
Rather like the meals in the Inns of Court, the effect of informal interaction is to socialise lawyers to be part of a collective effort. This is clearly an important feature of supranational courts such as the Court of Justice of the European Union, which organises regular visits from leading judges and others from its different legal systems, so they can understand the approach of the Court and that their concerns can be raised informally. Although this is important and care is put into the various forms of socialisation, it is difficult to demonstrate clearly any effect. The socialisation serves to provide both familiarity with the other institution and also personal contacts that serve as informal ways of understanding the other court.
There are many examples of what I would call "tracking". The national court is tracking the interpretations of supranational law made by supranational courts and implementing them (often as to the spirit and not just as to the letter). For example, the file prepared for major decisions by the Centre de documentation in the Conseil d'Etat or by the secretariat of the Conseil constitutionnel will provide background information on the case law of the European courts.  The dossier produced is an indication of what the court is considering. This is also shown by the contents of the arguments of the rapporteur public or the avocat général in cases (which are often now on line), and by the thematic dossiers which the courts produce summarising their own case law, but contextualising it in the light of the work of the European courts and other national supreme courts.  Mak's interviews with French supreme court judges confirms the importance these days of the decisions of the European Court of Human Rights, even for the Conseil constitutionnel.  As Vice-President Sauvé of the Conseil d'Etat remarked, the French courts will seek to avoid capitulation or rebellion in their engagement with the European courts . They have to decide the law as they see it, but influenced by respect for the expertise and authority of the European Courts. Similarly, there is mutual respect between the French supreme courts for their institutional position and their expertise.
The supranational courts also have their ways of keeping track on major developments within the national courts, notably through the cabinets of the judges and the advocates general.
4. Dialogue through judgments
Although Paterson is right to point to the way in which there is not much "dialogue" through the writing of judgments, there are important parts of the process which make courts open to engaging with each other. In a way, it is surprising that Paterson gained the impression that the CJEU was less open to dialogue than the Strasbourg Court. Paragraph 24 of the Luxembourg Court's Recommendations to national courts and tribunals in relation to the initiation of preliminary ruling proceedings  states:
If it considers itself able to do so, the referring court or tribunal may, finally, briefly state its view on the answer to be given to the questions referred for a preliminary ruling.
The preceding paragraph also encourages the national court to provide an explanation of the reasons prompting a reference, especially the relationship between provisions of EU law and those of national law. The CJEU also invites the communication of action taken as a result of its ruling and the eventual decision in the national proceedings. It may be that there a few cases from the highest courts: in the years of the UK's membership to the end of 2013, only 8% of UK references came from the highest courts and within those only 5 from the Supreme Court.  Such communications within the reference are more obvious forms of dialogue.
Bobek illustrates the issue of dialogue by referring to Melki and Abdeli.  In that case the Cour de cassation referred a question to the Luxembourg Court by way of a preliminary reference concerning its competence to make a reference. It submitted its own interpretation of national law, namely that national law precluded it from referring an issue of the legality of a national statute to Luxembourg before it had referred the issue to the Conseil constitutionnel. The French Government made different observations based on subsequent decisions of the Conseil constitutionnel and the Conseil d'Etat which gave a different interpretation of the national law, one which was more compatible with the requirements of EU law.  What we see here is the place of Government observations as a way of bringing together communications between different courts on the same topic. The communication is indirect, but it takes place.
Evidence for dialogue
Dutheillet de la Motte, a member of the Conseil constitutionnel, wrote that, although the Conseil does not cite the Strasbourg Court in its decisions, there are at least four ways in which the Conseil actually is influenced by the Court's case law.  In the first place, there are values which are clearly derived from its case law. An example would be the right to family life in article 8. The Conseil formally drew on the tenth paragraph of the Preamble to the 1946 Constitution ("The nation shall ensure…to the family the conditions necessary for their development"),  but it was clear that it was drawing on the Convention which had been used as a basis for similar rights by the Conseil d'Etat. Secondly, he argues that the Convention is used to enrich the content of domestic norms. Thus the freedom of expression in the Declaration of the Rights of Man of 1789 talks of the freedom to express and to publish. But in the Handyside decision of 1976,  the Strasbourg Court emphasised that the importance of pluralism in opinions. This was then picked up by the Conseil Constitutionnel in its Press Law decision of 1986. The freedom to publish opinions in 1789 became a freedom to receive a plurality of opinions in 1986, thanks to the Strasbourg Court's case law. Thirdly, he notes that the Strasbourg Court has significantly affected criminal procedure, especially through its case-law on article 6. For example, the rights of the defence in criminal proceedings is not well specified in article 16 of the Declaration of 1789, but this was interpreted in Décision n° 89-260 DC du 28 juillet 1989, recueil page 71 as implying "a just and fair procedure guaranteeing equality between the parties. The "equality of arms" case law is found in the work of the Strasbourg Court.  Finally, he notes the way in which the Conseil has actually been forced to change its own case-law notably in the area of retrospective validating legislation. Where the administration makes a legal error which is quashed by the courts, the legislator may seek to validate the acts retrospectively. The Conseil was happy to allow this as constitutional. For example, it accepted the validity of legislation of 1994 validating social security contributions. But this was reversed and in 1999, following case-law of the Strasbourg Court, it decided that it was up to the Conseil constitutionnel to review the legislator's justifications to see if they were sufficiently serious to justify removing a person's property.
The concept of "dialogue" has been used by a sociology professor who was also a judge of the Conseil constitutionnel, Dominique Schnapper.  In her view, there is a sense of rivalry between the different French national supreme courts, and that does seem right. The Cour de cassation and the Conseil d'Etat are different senior public bodies and have a pride in their own work. At the same time, they do follow the work of the other and consider whether the solution adopted is appropriate. In different areas, there are "lead" courts which others follow. Thus, in electoral matters, the Conseil constitutionnel (on national elections) follows the lead of the Conseil d'Etat (which deals with local elections and whose members are seconded to the Conseil constitutionnel to deal with legal issues arising in national elections).  In many employment matters, the legislator and the civil courts provide the standard for the administrative courts, which helps to lead to a convergence of case-law.  Mak  argues from her own interviews with judges in the French supreme courts that the role of the two European courts has triggered the development of connections between the national supreme courts to ensure that they have a coordinated position in relation to European norms. This may often involve informal mechanisms. For example, the Conseil constitutionnel does not have the time to refer pre-enactment matters to the CJEU, because it must decide cases within 30 days (8 days if there is urgency). But contacts within the court that come from meetings or from membership of the same corps of judges (e.g. the Conseil d'Etat) will enable informal contact to be established and an understanding gained by the French court of the European court's likely approach to a question.
Secularism ("laïcité") is a highly contested concept in France, but one which is seen as central to "republican" identity. Although there was much debate at the turn of the twentieth century, leading up to the explusion of Catholic religious orders in 1905, the Constitution of 1958 laconically says that France is a "secular" republic without further definition, and the only provision on fundamental rights to mention religion simply declares the principle of non-discrimination. In 1989 the Conseil d'Etat was called upon to provide an advisory opinion on the wearing of visible signs of belonging to a religion in public establishments, such as the wearing of the veil by Moslem women and girls in state schools, and then to make judicial rulings on the same question. The controversy did not go away and a Commission was set up under the Ombudsman, Bernard Stasi, which reported in 2003. Pages 10 to 18 of the Report explain that the common destiny of France comes through the "yeast of secularism". Following the Report, the Government proposed a bill which would outlaw the ostentatious wearing of religious signs in schools. This was passed very rapidly within three months of the Report. During the parliamentary deliberations, Jean-Paul Costa, then a judge of the European Court of Human Rights, gave evidence. Having been a senior member of the Conseil d'Etat when it produced its opinion in 1989, it was not surprising that he suggested that the ban on religious signs did not violate fundamental freedoms. He also expressed the opinion that, should the legislative rule come to his current Court, it would be upheld. The law was duly voted. In the June of the same year, a Section of the Strasbourg Court heard a case involving a ban on the wearing of Moslem dress in a Turkish university . The Court drew on the Stasi report and the French law as indicating a diversity of opinion in Europe and favoured a large margin of appreciation for national legislators. This view was then upheld in 2005 by the Grand Chamber (in which Costa sat as a Judge). The Strasbourg Court's ruling was also useful to the Conseil constitutionnel which, in November 2004 , had to judge the compatibility of the Constitutional Treaty of the European Union with the Constitution. In particular, it examined whether making the Charter legally enforceable. In particular, Art. II-70 of this Treaty protected religious freedom. The Praesidium of the Convention creating the European Constitution had declared that the Charter would be interpreted in the light of national constitutional traditions. This kind of deference to national authorities was seen also in the wide margin of appreciation accorded by the Section. It allowed a Turkish ban on the veil in universities because this fitted the Turkish secular tradition. The Conseil cited the Strasbourg Court's opinion explicitly to show that the French view on secularism would be permitted "which prohibits anyone from relying on religious beliefs to exempt themselves from general rules governing the relationships between public authorities and individuals." 
In this illustration, different supreme courts are puzzling over the same problem. As Government legal adviser, it is not surprising that the Conseil d'Etat was asked to take the lead. It would vet bills before they were presented to Parliament, and so broader guidance which was then incorporated into a ministerial circular to schools was desirable. That advisory opinion remained the benchmark for the internal debate within the French courts, even if a more general public inquiry was set up a few years later. A thoroughly considered legal opinion at national level was also influential in Strasbourg. No doubt supported by informal contacts with French judges within the Court, the Strasbourg Court adopted a fairly deferential position towards this considered national opinion. The Conseil constitutionnel also took its lead from the Conseil d'Etat. Formerly, the courts are not strictly bound by each other, and in a hierarchical system the more generalist court would not prevail over the specialist human rights courts. There are forms of communication here, a kind of informal dialogue.
This paper aimed to demonstrate that, to understand how France survives with five supreme courts, one needs to understand the informal processes in operation. French courts are not coordinated by giving one court the final say, an ability to impose a solution on the others. All the same, the system operates fairly effectively. The French citizen or the lower level French judge is not faced regularly with incompatible statements of what the law requires. It may happen at some times, but more typically there is sufficient harmony between the courts. This arises as a result of deference and understanding generated by informal processes.
In the first place, there is deference because the division of French supreme courts arises out of specialisation. One specialist court respects the expertise of another, and expects similar respect in turn. All the five supreme courts have a broad remit, but look at their problems from distinct perspectives. The professional character of each court encourages respect for the specialist remit of the others. Of course, there may be disagreements in areas of overlap, but most of these will not persist. At the very least, they will not pose a problem. For example, the French Conseil constitutionnel would uphold the priority of the French Constitution over either EU law or the Convention. Each of the latter two is law in France only because of the Constitution, rather than being above the Constitution. Whilst the Luxembourg and Strasbourg courts would claim their treaties have changed the legal scene in a radical way, they have not generated conflicts with the French courts.
The important part of the process is the understanding that one court has of the others. In part, this results from the flow of information and in part it is the result of the flow of personnel. The ability of one court to have a fairly accurate picture of the current positions of other courts is a prerequisite to informed decision-making, but also to avoiding accidental clashes of view. The existence of structures through which appropriate briefing is provided is essential. To go back to Paterson's work on the Supreme Court, he notes the importance of judicial assistants and also counsel in providing information. But the French courts have more systematic and better resourced systems through which information will come, and the same is true of the European courts. The other side is the flow of personnel from the judiciary, both at junior level into the Registry at Strasbourg or as référendaires in Luxembourg and at senior level. The mid-career appointment of British judges and the way careers develop at the Bar has the result that judges (or even potential judges) will not take early career secondments to Luxembourg or Strasbourg. At senior level, the UK has never appointed a sitting judge to the Strasbourg Court and no Strasbourg judge has ever returned to take up a judicial post in the UK. In the case of Luxembourg, a number of sitting judges have gone as judge or Advocate General from UK courts. But only one, Lord Slynn, has ever returned to take up a judicial appointment in the UK. Others have simply retired after their stay in Luxembourg. Paterson's researches obviously suggest that the greater closeness of the Luxembourg judiciary to judges in the highest UK courts does not lead to better dialogue. It may be that there is something distinctive about the UK position which is worth exploring.
Paterson's research encourages us to look at networking and flows of influence within a court and not just look at formal structures for securing agreement or alignment. The study of France which is sketched here (and which needs further work) would suggest that Paterson has identified the right tools for the study of the judiciaries in other countries, though this may lead to the investigation of different factual issues in each country.
 Professor of Law; Director of the Centre for Public Law, University of Cambrdige.
 A. Paterson, Final Judgment: the last law lords and the Supreme Court (Cambridge, Cambridge UP 2013), 223.
 Ibid., p. 224.
 M. Lasser, Judicial Deliberations (Oxford).
 On the French court system summarised here, see J. Bell, S. Boyron & S. Whittaker, Principles of French Law (2nd edn, Oxford, Oxford UP 2008), ch. 2.
 J. Bell, French Constitutional Law (Oxford, Oxford University Press 1992), 53-4; J. Bell, French Legal Cultures (London, Butterworths, 2001), 217-9, 237-9.
 S. Wright, 'The Conseil constituionnel since Mid-2007' (2010) 16 European Public Law 539.
 J. Bell, French Legal Cultures, 95-7, 192-5
 S. Braconnier, Jurisprudence de la Cour européenne des droits de l'homme et droit administrative français (Brussels, Bruylant 1997), 36. He cites Cass.crim 4 May 1994, Saïdi, JCP 1994 II 22349 "decisions handed down by the European Court of Human Rights under the terms of articles 19 and 50 to 54 of the Convention have no direct incidence in internal law on the decisions of national courts".
 Jean-Paul Costa, La Cour européenne des droits de l'homme. Des juges pour la liberté (Paris, Dalloz, 2013), 148-52 and 178-82.
 Above note 8, p. 515.
 M. de S.-O.l'E. Lasser, Judicial Transformations (Oxford, Oxford University Press 2009), 279.
 See J. Bell. 'The Expansion of Judicial Review over Discretionary Powers in France'  Public Law 99,112: CE 9 June 1978, Sieur Lebon, AJDA 1978, 573, conclusions Genevois.
 Bell, French Constitutional Law, ch.2.
 CE 21 June 1895, Cames, D. P. 1896.3.65, concl. Romieu.
 Cass. 16 June 1896, Teffaine, D.P. 1897.1.433.
 See for example, the discussion of Cass. 2 civ., 28 March 2013 in the Cour de cassation, Rapport Annuel 2013 (Paris, Documentation française 2014), 587.
 See art. 5 of the Civil Code of 1804; S. Belaid, Le pouvoir créateur
 See Lasser, Judicial transformations, ch.7.
 Ibid., pp. 183, 197.
 See J. Bell, French Legal Cultures (London, Butterworths 2001), 61-63; and J. Bell, 'Interpretative Resistance faced with the Case-Law of the Strasbourg Court' (2008) 14 European Public Law 137.
 See CE Ass. 8 December 1978, GISTI, Leb. 493; Braconnier, pp. 98-101.
 For a presentation of the kinds of overlap that exist between Conseil d'Etat judges and other courts and social institutions in France, see B. Latour, La fabrique du droit (Paris, La Découverte 2002), 137.
 See Costa, pp. 196-8.
 E. Mak, Judicial Decision-making in a Globalised World (Oxford, Hart 2013) (hereafter "Mak"), 90-94.
 See Costa, pp. 223-5.
 See http://www.coe.int/t/DGHL/cooperation/ccje/default_en.asp.
 M Claes and M de Visser, 'Are You Networked Yet? On Dialogues and European Judicial Networks' (2012) 8 Utrecht L. Rev. 100, 106-112.
 Ibid, 111.
 Ibid, 111-12.
 Ibid, 113.
 Paterson, op.cit.
 Mak, above n 25.
 Mak, see especially pp.104-5,113.
 Operating the principle da mihi factum, dabo tibi ius, the judge is responsible for finding the relevant legal authorities and this is often supported by a central research unit in a large court.
 For example, the dossier on the Conseil d'Etat website on "Le juge administrative et l'expression des convictions religieuses".
 Mak, pp. 171-73.
 Mak p. 75.
 OJ 2012 C338/01.
 Court of Justice, Annual Report 2013, p. 109. This compares with 33% of references from Germany and 21% from France.
 Joined cases C-188/10 and C-189/10, Melki and Abdeli  ECR I-5667: M. Bobek, "Of Feasibility and Silent Elephants: The Legitimacy of the Court of Justice through the Eyes of National Courts" in M. Adams, H de Waele, J. Meeusen & G. Staetmans, Judging Europe's Judges (Oxford, Hart 2013), 197, 225.
 See paras. 33-36, 46-9.
 O. Dutheillet de La Motte, 'L'influence de la Cour européenne des droits de l'Homme', 13 February 2009 : http://www.conseil-constitutionnel.fr/conseil-constitutionnel/root/bank_mm/pdf/Conseil/cedh_130209_odutheillet.pdf (last visited 24 January 2015).
 Décision n° 93-325 DC of 13 August 1993, Recueil 224.
 Handyside v UK , Application no. 5493/72  ECHR 5.
 See Golder v UK, Application no. 4451/70  1 EHRR 524
 Une sociologue au Conseil constitutionnel (Paris, Dalloz, 2010).
 Mak, p. 75.
 See A. Morin-Galvin, La Convergence des Jurisprudences de la Cour de cassation et du Conseil d'Etat (Paris, LGDJ, 2013)
 Mak, p. 74.
 Application no. 4774/98 of 29 June 2004.
 CC decision no. 2004-505 DC of 19 November 2004, para. 18: see generally J. Bell, 'French Constitutional Council and European Law' (2005) 54 ICLQ 735.
 This formulation echoes p. 17 of the Stasi Report: "secularism balances recognition of one's own identity with the efforts necessary to knit together individual beliefs with social bonding.