1 - Introduction

Summary

This chapter argues for the need to revise the features seen as distinctive by British scholars who have studied French administrative law in the past. There is a distinctive path dependence arising particularly from the Napoleonic reforms and from the separation of public and private law courts. But modern French administrative law is shaped by the rise of French constitutional law, EU law and the European Convention on Human Rights. The influence of these is illustrated. Contemporary trends in French public administration are also discussed.

1.1 French Administrative Law in British Scholarship

French droit administratif has been a subject of fascination for British lawyers since the late nineteenth century. Although, on first reading, Dicey’s An Introduction to the Study of the Law of the Constitution, first published in 1885, might appear to have rejected droit administratif as contrary to the British understanding of the ‘rule of law’,¹ John Allison has admirably demonstrated that Dicey’s subsequent writing (often unpublished) reveals a detailed understanding of and admiration for the achievement of droit administratif.² In those writings, he explains that it is a misconception not to consider droit administratif as law.³ He also explains how French administrative judges have become not just officials who judge cases, but almost equivalent to judges.⁴ His particular concern remained that relations between the citizen and the state were governed by different principles to private law relations between citizens and that adjudication was not determined in the ordinary courts.⁵ These were key tenets of the British conception of the rule of law which differed from the French and which excluded the existence of administrative law in England.

The published (and less subtle) views Dicey expressed reverberated for most of the following century. Later generations of scholars who sought to establish administrative law as a subject used French droit administratif as a positive benchmark of what the common law could achieve. Port in 1929 was one of the first writers of a treatise on ‘administrative law’ in the UK.⁶ Port described French administrative law, discussing the theories of Hauriou, Jèze and Duguit.⁷ He then used French categories to describe American administrative law. Even if neither he nor the other main UK writer on administrative law at the time, Robson,⁸ subscribed to Dicey’s approach to administrative law, they retained his idea that France was the primary reference point for conceptual ideas, a point supported by the content of contemporary journal articles and by the contributions of Robson and Laski to the Donoughmore Committee.⁹ This continued after the Second World War with the work particularly of Hamson in his Hamlyn lectures in 1954¹⁰ and of J. D. B. Mitchell in Scotland.¹¹ But it would be fair to say that the apogee of French droit administratif as the benchmark of a developed administrative law was reached in 1956 when the Vice President of the Conseil d’Etat was invited to give evidence to the Franks inquiry into the control of ministers’ powers. But that committee did not choose to recommend any features of the French model.¹² The American model, especially as it developed with the Administrative Law Procedure Act 1945, became too alluring for the common lawyer.¹³ Nevertheless, the publication of a textbook on French administrative law by Neville Brown and Jack Garner provided the English-speaking lawyer with a good insight into droit administratif just as administrative law was beginning to take shape in Britain.¹⁴

The theme of the twentieth-century works on French droit administratif were both that France had a sophisticated and effective set of legal principles to review the exercise of power by the executive and that it was distinctively French in terms of its organisation and sources. No doubt this theme was encouraged by the talks and writings of members of the Conseil d’Etat and French academics. Indeed, that distinctiveness may well have been the reason why the American model was more attractive to the British common lawyers (apart from the linguistic accessibility of its judicial decisions and scholarly writings).

The theme of this book is different. In the long period since Hamson, Brown and Garner wrote their works, France has changed, and French administrative law has changed. The most important change has been the active participation of France in the European Union (EU) and in the Council of Europe with its European Convention on Human Rights. France helped to found the European Coal and Steel Community in 1951, and it ratified the Treaty of Rome founding the European Economic Community (EEC) in 1957, which was opposed by the Gaullists who came to power in 1958 and created the Fifth Republic. Arguably, France was not reconciled to the EEC until at least De Gaulle’s abdication of power in 1969, if not until the election of Giscard d’Estaing as President in 1974. France did not ratify the European Convention of 1950 until 1974 and did not allow direct petition until 1981. But once France did ratify these treaties, the primacy given to treaties over national legislation under the Constitutions of the Fourth and Fifth Republics gave a strong impetus to the influence of these agreements on subsequent French domestic law, including administrative law.

The relationship between French administrative law and principles of EU or European Convention law has not always been easy. Two topics illustrate this point: the recognition of the supremacy of EU law over national law and the right to a fair trial as it affected long-established procedures in the administrative courts. These topics will be discussed in some depth in Sections 5 and 6 of this chapter.

1.2 What Is ‘Droit administratif’?

In many ways, Dicey understood droit administratif very well. He wrote:

Droit administratif, as it exists in France, is not the sum of the powers possessed, or of the functions discharged by the administration, it is rather the sum of the principles which govern the relationship between French citizens, as individuals, and the administration as the representative of the state.¹⁵

There are clearly two dimensions. On the one hand there is the internal dimension of administrative law as the principles which govern the division of tasks within the administration, whether this be civil service employment, or the power to delegate functions, or the supervision of the functions of specific administrations by a ministry or a prefect. On the other hand, there are the external relations of the administration towards citizens (or, as the French more correctly call them, ‘the administered’). As Dicey rightly saw, the French believe that the relations between the citizen and the state should be governed by different principles from those governing relations between citizens. The state is acting in the public interest and so is given special powers to achieve that objective, whereas private citizens act in their own interest and have less justification for interfering with the interests of others.

So the distinctiveness of droit administratif does not lie in the distinctive character of the judges, their formation and careers (which will be seen in Chapter 3). Nor does it lie in the procedure which has been aligned increasingly to that in private law and in other European Convention countries (as will be seen in Chapter 4). Instead, the distinctiveness lies in the powers and responsibilities which attach to the state in its relationship with the citizen. The mission to fulfil the general interest (l’intérêt général, as the French put it) confers on the state extraordinary powers (pouvoirs exorbitants) which no citizen could exercise over another – for example, expropriating the property of an individual to build a new TGV line. Furthermore, unlike the private individual, the state has the authority to act without consent (un pouvoir unilatéral) – for example, to impose a curfew or to terminate a contract. On the other hand, the state has special responsibilities. The first is that it has to justify its actions in a way a private individual does not. The state has to show that its actions are authorised (the issue of compétence), that they will lead to a permitted objective, and that they are not excessive in the burdens they impose (absence of mesure excessive: see Chapter 7, Sections 1.5.3 and 1.5.4). A private individual can buy a house on a whim and need not give reasons why they do not wish to continue negotiations (unless the conduct is in very bad faith). The state cannot act on a whim because that would be an abuse of power (un détournement de pouvoir). It has to act for lawful reasons and, these days, it has to provide those reasons to the person affected. As a result, distinctive rules apply to public procurement that do not apply to private procurement (see Chapter 9). Despite the emphasis of Dicey, the state is not subject to the same rules of contract as private individuals. Furthermore, the rules of competition that apply to private individuals are weaker. The private individual is required not to make agreements with others that distort competition and not to abuse a dominant position. The state is almost assumed to be occupying a dominant position and is strictly controlled in the way it chooses its contracting partner. As regards liability to private citizens, the Revolution recognised the principle of the equality of public burdens in art. 13 of the Declaration of the Rights of Man, and so where one citizen suffers an excessive detriment from a policy, then the state has to compensate them (see Chapter 8). This is different from a private individual who normally only has to pay compensation for a wrongful harm. The state also has to pay when it has done a wrong. But the fault of the public service extends to a failure to deliver the service which should be expected by the user – for example, the failure to provide lessons in particular subjects at school. This would be treated in England more as maladministration than fault. This is in addition to liability to compensate citizens for excessive detriments suffered as a result of (lawful) public policies. The distinction between public and private law is difficult to make in some instances (see Chapter 5), but the overarching idea that Dicey spotted is that the state is not just one subject of the law like any other subject of the law. In the French sense, the state under law (l’Etat de droit or le Règne du droit) means that the actions of the state are governed and controlled by law. But, unlike Dicey’s conception of the rule of law, that does not mean that the state or its officials are subject to the same rules as the private individual. The scrutiny of whether an act is lawful is more stringent, and the rules of liability to compensate are more extensive.

1.3 The Shaping of Droit administratif

As will be seen in Chapter 2, the general principles of droit administratif – the review of administrative decisions, liability in contract and extra-contractually, and administrative procedure were not codified at the same time as private and criminal law were in the Napoleonic period. As a result, droit administratif was largely the creation of the administrative judges, who were, for the first 150 years, just the members of the Conseil d’Etat. They shaped the subject not only through judgments, but also through the arguments of the commissaire du gouvernement (now called the rapporteur public) and through the textbooks and scholarly articles which individual members wrote extrajudicially. In that way, it was more like the common law in which case law, rather than statute, has set out major principles for judicial review of administrative action, contract and tort. In many ways, one of the high points of this process of developing administrative law occurred just after the Liberation in 1944. In the absence of binding legal statements of fundamental rights, the Conseil d’Etat developed a set of legally binding ‘general principles of law’ which bound the administration, even if they could not limit the sovereignty of the legislature, except by way of interpretation (see Chapter 2, Section 6.3). These principles consolidated the understandings of democratic liberal principles as developed in the Third Republic (1870–1940) and taken forward in the Fourth Republic (1946–58).

The full importance of the administrative judiciary and scholarly writers in shaping the French droit administratif will be explored in Chapter 2. But it is important to understand that contemporary French public law is shaped not only by the administrative judges and scholars. Since 1958, three sources of influence have emerged which are very significant in shaping the general principles and sometimes the rules that govern the relationship between the state and those it administers. The first is purely internal – the Constitution and the Conseil constitutionnel, which has emerged as a constitutional court. The second and third are shared with other European countries, but also have a direct influence on domestic administrative law through provisions within the 1958 Constitution – membership of the European Union and the ratification of the European Convention on Human Rights. Certainly, until after the second edition of Brown and Garner was published in 1973,¹⁶ the Conseil d’Etat with its droit administratif was supreme in shaping public law in general and the law relating to the administration in particular. But since the 1970s, first the Conseil constitutionnel, then the Court of Justice of the European Union (CJEU) in Luxembourg and the Court of Human Rights in Strasbourg have exercised influence over the general principles of administrative law, and sometimes over its detail. We therefore need to be aware of these other factors which create the climate in which droit administratif now operates.

1.4 The Influence of French Constitutional Law

In Chapter 2, we will look in more detail at the place of constitutional law among the sources of French law in general and of administrative law in particular. The Constitution of the Fifth Republic in 1958 did not set out any new constitutional principles relating to fundamental rights and did not create a constitutional court. It is quite clear that the provisions on fundamental rights mentioned in the Preamble to the 1958 Constitution were originally intended to be conventional, not legal. When asked specifically whether the provisions of the Preamble were to be of constitutional value, the commissaire du gouvernement, Janot, replied, ‘Certainly not!’ They were to be binding on the Government, but not on Parliament.¹⁷ In other words, they would be legally enforceable on the Government by the administrative courts, but only politically enforceable on Parliament, a solution which some found unacceptable.¹⁸ As art. 5 of the Constitution made clear, the President of the Republic, not the Conseil constitutionnel, was to be the guardian of the Constitution, much as had been the role of the President in the Third and Fourth Republics. On this view, the President is not amenable to legal sanction for his interpretations of the Constitution, but these therefore fall into the area of conventional constitutional obligations, rather than legal obligations.¹⁹ This initial understanding of the Constitution has changed radically. It is very clear that it contains legally binding principles which affect the administration.

In the middle of the twentieth century, there was a dispute between two of the titans of French public law at the time, Vedel and Eisenmann, as to whether there were constitutional foundations.²⁰ Vedel argued that ‘the Constitution is the necessary foundation of the rules which together make up droit administratif’.²¹ The actions of the executive in exercising special public powers that exceeded those of private individuals was acknowledged in certain constitutional texts, especially those of the Fourth and Fifth Republics, and in certain case law of the Third Republic. The texts conferring powers on the President of the Republic and on the Prime Minister presupposed that there was a domain in which the executive exercised sole competence. Furthermore, the 1958 Constitution specifically gave autonomous legislative powers to the executive in art. 37. This seemed to confirm the areas of sole administrative competence. Eisenmann took the view that no constitutional text clearly set out the powers of the executive. In any case, the consequence of administrative law being grounded in the Constitution would be that it could vary from one constitution to another, whereas the experience of the Conseil d’Etat was of a continuity of administrative law principles despite changes in the Constitution, particularly in 1946 and 1958. He saw administrative law as grounded in the sovereignty of Parliament. By that he meant that the powers the Constitution granted to the state were to execute the laws enacted by Parliament and the courts had the function of giving the correct interpretation of the powers given to the state. In essence, Vedel was keen to argue that the Constitution conferred a special position on the state to exercise extraordinary and unilateral powers to fulfil its mission. This included legislative powers, as is shown by art. 37 of the Constitution and by the First World War case law of the Conseil d’Etat on the inherent powers of the President to maintain public order and to manage the public service.²² (We see here echoes of the discussion in the UK of the nature of the prerogative over the civil service in CCSU v Minister for the Civil Service.²³) On the other hand, Eisenmann argued that the scope of executive action depended on what Parliament authorised the executive to do.

To an important extent, Vedel had the final word, not as a scholar, but as a judge of the Conseil constitutionnel. As a former President of the Section du Contentieux of the Conseil d’Etat and himself a leading administrative law scholar, Bernard Stirn, remarked ‘through its case law, the Conseil constitutionnel has enriched “the constitutional sources of administrative law”’.²⁴ In the period up to 1970, the Conseil d’Etat had been central in shaping the protection of fundamental rights through its notion of ‘general principles of law’, often based on the Declaration of the Rights of Man of 1789, itself not seen at the time as having legally binding status. But it refused explicitly to challenge the legality of legislation. Vedel was reporting judge for two key decisions of the Conseil constitutionnel which gave constitutional force to key principles of administrative law. In 1980,²⁵ the Conseil constitutionnel endorsed the independence of the administrative courts as a fundamental principle recognised by the laws of the Republic. In 1987, it found that the judicial review of decisions of bodies exercising executive power belonged to the administrative courts, thereby consecrating the separation of administrative and ordinary courts by way of a fundamental principle recognised by the laws of the Republic (even if the best statement was in a law of the Bourbon monarchy in 1790).²⁶ So a law conferring such powers on the ordinary courts was unconstitutional. Rather than focusing on the rules concerning the powers of the administration, these decisions focus on the control of administrative powers, appealing to the idea of the separation of powers, rather than the rule of law (as UK courts would have done). It is the control of the administration that was the object of constitutional attention in 1641, 1790, 1799, 1872 and 1945, albeit not all in texts that are these days considered legally binding.

But, as Stirn pointed out,²⁷ particularly in the past decade or so, there is a spirit of cooperation between the Conseil constitutionnel and the Conseil d’Etat in developing the constitutional principles that underpin droit administratif. The reform of the Constitution in 2008 created the possibility for the first time that laws which had already been enacted could be challenged for unconstitutionality. Previously, the Conseil constitutionnel was only concerned with laws before they were promulgated. Now it is possible for a litigant in a civil or administrative case to challenge the effect of a law on the ground that it is unconstitutional. In this process, the top court in each system acts as the gatekeeper to ensure only serious issues are submitted to the Conseil constitutionnel. The Conseil constitutionnel deals with the constitutional question by way of a reference from the administrative or ordinary courts – hence it is called a preliminary question, the question préalable de constitutionnalité (QPC). This innovation has changed the role of the Conseil constitutionnel. In the years since 1 March 2010, when the QPC came into force, the Conseil constitutionnel has typically dealt with references from parliamentarians on between twenty-five and thirty laws a year prior to promulgation, but between seventy and eighty QPC references. Of the references received in 2019, 46 per cent were from the Conseil d’Etat. The Conseil d’Etat can act as gatekeeper. For example, in 2018, it refused to submit a law on terrorism to the Conseil constitutionnel because it did not think the complaint of unconstitutionality was sufficiently serious. In its view, the legislator had provided sufficient safeguards for fundamental rights that no breach of constitutional values was arguable.²⁸ On the other hand, the decisions of the Conseil constitutionnel on a QPC reference can lead to changes in the way the administration or the administrative courts work. A good example will be seen in Chapter 4, Section 2.7, on the composition of specialised administrative courts (what the UK knows as tribunals). There the decision of the Conseil constitutionnel led to a restructuring of the membership of these bodies and the transfer of much of their work to the generalist administrative courts.

The constitutional principles requiring a hearing before a sanction is imposed is recognised both by the Conseil d’Etat and by the Conseil constitutionnel.²⁹ That affects the way the administration behaves, as well as how the legislature drafts the powers it confers on the administration.

1.5 The Influence of EU Law: French Administrative Law and the Supremacy of EU Law

Entry into the European Economic Community (as it then was called) in 1957 was politically divisive in France. Both the communists and the Gaullists were against it. De Gaulle, who returned to power in 1958, blocked much activity through his ‘empty chair’ policy in 1965 at a time when the EEC had to act by unanimity. It is therefore not surprising that the Conseil d’Etat did not accept the supremacy of EEC law over domestic law when the issue was raised before it in 1968. In Semoules de France, there was a clear conflict between an EEC regulation and a French Law.³⁰ The Conseil held that it had no power to ignore a constitutionally valid law, and so it refused to give effect to the EEC regulation because the law was posterior to the regulation and therefore expressed the last will of the sovereign Parliament, despite art. 55 of the Constitution according to which a regularly adopted treaty must prevail over a law. When a similar issue returned ten years later, the response was much the same with regard to the effect of a directive towards an administrative act. In Cohn-Bendit, a German leader, brought up in France, of the May 1968 student protests was subject to an expulsion order imposed in 1968.³¹ In 1976, he requested the Minister of the Interior to revoke the expulsion in conformity with art. 6 of an EEC directive of 1964. On the Minister’s refusal without reasons (as required by the 1964 directive), Cohn-Bendit challenged the decision and the tribunal administratif referred the matter to the European Court of Justice (ECJ). The Minister appealed successfully to the Conseil d’Etat. Despite the clear ruling of the ECJ on the direct effect of art. 6 of the 1964 directive in relation to France,³² the Conseil d’Etat held that, even if Member States were obliged to legislate to implement a directive, ‘these authorities alone remain competent to decide the form in which to implement directive and to determine themselves … the appropriate means to give effect to it in internal law’. This contrasted with approach of the Cour de cassation, which had already decided in 1975 to give priority to EEC law over a conflicting national law.³³ To be fair to the Conseil d’Etat, it was not the only national court which took this approach to the idea that directives might be directly effective – indeed in 1981, the Bundesfinanzhof expressly agreed with the Conseil. But there was clearly a generational problem among the judges in attitudes towards EU law. Indeed, the commissaire du gouvernement Bruno Genevois had taken the opposite view with his famous claim that ‘At the European community level, there should be neither a judges’ government nor a judges’ war, there must be room for a dialogue between judges.’ In 1988, the Conseil constitutionnel, sitting as an election court, held that EEC law had to prevail over an inconsistent national law.³⁴ In two decisions in 1989, the Conseil d’Etat followed suit. In Compagnie Alitalia, it held that there was a principle under which an administrative authority was obliged to withdraw an illegal decision at the request of its addressee (abrogation).³⁵ This principle applied not only to decisions contrary to national law, but also to those contrary to European law. In this case, the Minister was obliged to withdraw national regulations in the General Tax Code which were inconsistent with the sixth EEC VAT directive. This was followed by Nicolo in which an individual challenged a 1977 French law on European elections on the ground that it gave voting rights to citizens of France’s overseas départements and territories on the ground that this was contrary to art. 227–1 of the EEC Treaty.³⁶ The Conseil held that the 1977 law was perfectly compatible with art. 227–1, but in so doing, it recognised (albeit obliquely) the superiority of EEC law, a point made clear in the conclusions of the commissaire du gouvernement Frydman.

The Conseil’s slowness to recognise the supremacy of EEC law (as it then was) was matched by that of the House of Lords, which came to a similar decision in the same year in the Factortame decision.³⁷ The position of both courts took place against the background of the creation of the Single Market from 1986 to 1992, which gave a new impetus to the EEC. It reveals that courts at the time were reluctant to be at the forefront of greater transfers of sovereignty to the EEC, and that they relied on constitutional reform which came with the constitutional amendments of 1992, including art. 88–1 of the Constitution, which gives explicit priority to EU law (as it was called after the Maastricht Treaty of 1992).

Two decisions in the 2000s show the sincere adhesion of the Conseil d’Etat to EU law. In the Arcelor case of 2007, it had to deal with the issue of competition between a European rule, in this case a directive setting up the system of greenhouse gas quotas but only for certain industries, and the Constitution. Since plastic industries were not concerned by this new regulation, unlike steel industries, Arcelor challenged it on the ground the decree transposing the directive was contrary to the constitutional principle of equality. The Conseil d’Etat ruled that, since there was an equivalent principle at the EU level, it should decide on the basis of EU rather than on constitutional principle, which the commissaire du gouvernement Guyomar called an ‘opération de translation’.³⁸ The Conseil d’Etat asked the ECJ for a preliminary ruling, which eventually ruled that the breach of equality was justified on the ground that it was a complex system which must be put in place step by step.³⁹ However, it remains possible, in theory, that EU law may not offer an equivalent principle to a French constitutional one – one may think of secularism – which may lead to the Constitution to prevail over EU law.

In 2009, the Conseil d’Etat overturned the Cohn-Bendit case law, regarding the directive no 2000/78/CE not transposed at the time of the litigation, which imposes Member States to secure a reverse burden of proof whenever an individual presents in court facts from which it may be presumed that there has been direct or indirect discrimination.⁴⁰ The Conseil d’Etat applied this to a civil law judge who challenged her rejection as a professor in the Ecole Nationale de la Magistrature, which she claimed was based on her trade union membership. Although the Conseil d’Etat admitted that certain facts might raise a presumption of a potential discrimination, her claim was dismissed on the ground that the woman nominated instead of the claimant had better qualifications for the position, based on various evaluations of the two as well as the linguistic capacity of the nominee.

It is significant that, when faced with a French government claim that the CJEU had interpreted EU law in a way which was contrary to the French Constitution, the Conseil d’Etat did not rise to the bait, but sought to diffuse the problem by aligning EU law with domestic constitutional law. Its reaction was unlike that of the German Constitutional Court in Weiss.⁴¹ The decision of the Conseil d’Etat in La Quadrature du Net carefully negotiated the French policy of wishing to have access to mobile telephony data in the fight against terrorism with the EU legislation on data retention.⁴² Guided by a reference to the CJEU, it found lawful most of what the government wished to ensure for its antiterrorism policy, but required the retention to be reviewed more frequently than the government planned. In interpreting the French legislation, it ensured respect for the French constitutional objective of protecting public order and respect for privacy within EU data protection law. The Prime Minister argued that the requirements of the European Court of Justice in its reply to the reference from the Conseil d’Etat conflicted with the constitutional objectives of protecting public order and the investigation of crime. The decision of the CJEU had been that the EU directives on data protection ‘must be interpreted as precluding legislative measures which, for the purposes laid down in Article 15(1), provide, as a preventive measure, for the general and indiscriminate retention of traffic and location data’.⁴³ But the reply of the CJEU went on to state that general and indiscriminate requirements to hand over data would be allowed for the purposes of safeguarding national security, recourse to an instruction requiring providers of electronic communications services to retain, generally and indiscriminately, traffic and location data in situations where the Member State concerned is confronted with a serious threat to national security that is shown to be genuine and present or foreseeable, where the decision imposing such an instruction is subject to effective review, either by a court or by an independent administrative body whose decision is binding, the aim of that review being to verify that one of those situations exists and that the conditions and safeguards which must be laid down are observed, and where that instruction may be given only for a period that is limited in time to what is strictly necessary, but which may be extended if that threat persists.

The Conseil d’Etat found that the French situation fitted within the exception and so there was no clash between EU law and domestic constitutional requirements.

1.6 The Influence of the European Convention on Human Rights⁴⁴

Article 6 (1) of the European Convention on Human Rights provides ‘In the determination of his civil rights and obligations or of any criminal charge against him, everyone is entitled to a fair and public hearing within a reasonable time by an independent and impartial tribunal established by law.’ It is the concept of ‘an independent and impartial tribunal’ that caused the greatest difficulty for French administrative law in relation to the role of the commissaire du gouvernement.

The commissaire du gouvernement was a long-established part of French administrative court procedure.⁴⁵ Created in 1831, the commissaire du gouvernement had as his function to present arguments to the court in the interests of the law, and not in the interests of any of the parties, either the government or the citizen. He is conceived as part of the judicial function and he is a judge.⁴⁶ As the Conseil put it in 1957:⁴⁷

Considering that the commissaire du gouvernement in litigation before the Conseil is not the representative of the administration; … as his mission is to search out and present to the Conseil the issues to be resolved in each case and to make known his views, formulated completely independently, his assessment, which should be impartial, of the facts of the case and the applicable rules, as well as on the solutions which, according to his own opinion, are required.

His role is to advise the court neutrally and to maintain a degree of continuity within the case law of the court. An experienced commissaire described the role as one of ‘forging the case-law’ as well as publicising it to the wider world.⁴⁸ He was able to agree to the order of hearing of cases and to decide the importance of a case and request a hearing before a more solemn formation of the court. In this respect, as guardian of the case law, he performed a role close to that of the Advocate General before the European Court of Justice, which was modelled on the procedure in the French courts. The commissaire enjoyed independence in formulating his arguments and, despite the name, was never subjected to orders from the government.

But two features of the Conseil d’Etat procedure (and that of other countries which followed its procedure) attracted adverse comment from the European Court of Human Rights in the 1990s, years during which judicial independence was a particular concern, both in relation to the investigation of political corruption and in establishing democracy in Central and Eastern Europe following the fall of the Iron Curtain. The first of these was the connection between the advisory and the adjudicatory functions of members of the Conseil d’Etat. As will be seen in Chapter 3, the Conseil d’Etat has long had a function as legal advisor to the government, as well as of judge of governmental actions. In Procola v Luxembourg, the European Court of Human Rights ruled that judges who had advised the government on the legality of legislative or administrative instruments could not then adjudicate on cases involving their application.⁴⁹ The Luxembourg Conseil d’Etat had so few members that it was difficult to avoid a situation in which those who had advised on draft legislation would not also be needed to make up a judicial panel adjudicating on issues connected with the legality of the measure in question. The French took the view that this problem would not affect them, because there were far more members of the French body, so a separation of functions could be maintained.

A greater challenge to the French conception of judicial independence and of a fair trial came in a series of cases questioning the role of the commissaire du gouvernement. Although the function (like that of the avocat général in the ordinary courts) was to be the neutral advisor of the court, a number of procedural aspects of the role attracted criticism from the European Court of Human Rights. There were two main concerns. The first was that the conclusions presented by the commissaire du gouvernement were not open to challenge by the parties. The second was that the commissaire du gouvernement participated in the deliberations of the court. As we will see in Chapter 4, the commissaire (now called the rapporteur public) speaks after the parties have made their submissions.

In Kress v France, a claimant in a damages action against the state complained about a number of breaches of art. 6 of the European Convention in relation to the hearing of her case before the Conseil d’Etat.⁵⁰ The first was that she did not have access to the opinion of the commissaire du gouvernement. This was rejected. Before the administrative courts, the parties could make observations on his remarks by way of a short, written submission to the court, known as a note en délibéré. This was different from other cases in which European Court of Human Rights had sanctioned civil and criminal procedure in Belgium and France because the parties did not have access to the conclusions of the avocat général, but had no way of rebutting points before the judges deliberated.⁵¹ In administrative court procedure, the claimant’s avocat was permitted to ask the commissaire du gouvernement for an indication of the line of his arguments before the hearing and could send in a note before the judges deliberated. This made the procedure different from that of the Cour de cassation and saved this aspect of the procedure from censure. All the same, the European Court of Human Rights still declared the role of the commissaire du gouvernement incompatible with art. 6 of the Convention. It simply adopted a different perspective from the French courts based on the idea that justice must not only be done, but must be seen to be done. In essence, it considered the commissaire du gouvernement as an amicus curiae, someone offering impartial advice to the court, not part of the judicial team who decides the case. The analogy with the Advocate General in the European Court of Justice in Luxembourg was only too obvious.⁵²

The key problem with the role of the commissaire du gouvernement was that, in status, he was a judge and considered by the French as part of the judicial team dealing with the case. Indeed, the commissaire du gouvernement had actually read the full file and worked on the cases attached to it. Only the rapporteur in the judicial team had done as much work. As will be seen when discussing court procedure in Chapter 4, the other deciding judges will not necessarily have read the case file in anything like such depth. Furthermore, the commissaire had access to the draft judgment prepared by the rapporteur before the oral hearing. Indeed, one of the authors had the experience in the 1980s of reading case files on the day before the hearing and could compare the draft judgment of the rapporteur with the draft conclusions of the commissaire du gouvernement on the same cases. A particular problem was the practice of the administrative courts that the commissaire retired with the deciding judges and was present during their deliberations. He was allowed to speak, but not to vote. One of the authors was allowed to be present during the deliberation phase of cases in both the Conseil d’Etat and in some tribunaux administratifs during the 1980s and can vouch for the fact that the commissaire did indeed speak during the deliberations at the invitation of the deciding judges and there was often a debate with him.⁵³ This aspect of the commissaire’s role was considered unacceptable by the European Court of Human Rights. It relied on the theory of appearances, so beloved of the common law approach to natural justice. The Court thought that the litigant was entitled to be assured that the very presence of the commissaire du gouvernement could not exercise and influence on the outcome of the court’s deliberations. On a French analysis, this argument was stupid. If the commissaire du gouvernement was part of the judicial team, then of course he ought to be able to influence the decision, even if he does not have a vote. On the other hand, the European Court was fixated with an analysis of the judicial bench, more familiar in the common law, where there is a clear separation between advocates (representing the parties or the public interest) on the one hand, and judges sitting on the bench on the other. Only if the reasonable litigant’s perspective adopted such a strict separation and treated the commissaire as an advocate and not a judge did this appearance seem problematic. No one practising in the French system would adopt such a perspective. But the alternative view prevailed. The French administrative judicial establishment reacted badly to this.

Not surprisingly, the initial reaction of the French administrative judges was to stick as far as possible to their traditional practices and to make minimal changes – for example, by ensuring the parties were aware of the ‘sense’ (but not the detail) of the commissaire’s arguments in advance of the hearing. By contrast, the Cour de cassation had decided to comply with Strasbourg court ruling against it,⁵⁴ and no longer to allow the avocat général to be present at the private deliberation phase with the deciding judges. The position of the administrative courts came under further scrutiny from the European Court of Human Rights in Martinie v France.⁵⁵ Here the issue was the compatibility of the procedure before the Cour des comptes (the audit court judging public accounts and disciplining public accountants, see Chapter 3) with art. 6 of the Convention. The procedure of the Cour des comptes was similar to that of the general administrative courts, except that it was purely a written procedure. The majority of the Grand Chamber found that the procedure violated art. 6, despite a vigorous defence led by the French judge (who was also a member of the Conseil d’Etat). Two grounds of the decision were significant for all French administrative courts. France was condemned because the procureur général (the equivalent of the commissaire du gouvernement in the general administrative courts) was present during the deliberation, even though he did not in fact participate. In addition, the report of the reporter judge (i.e. the draft judgment) was communicated to the procureur before the hearing, but not to the parties, so he had privileged access. The principle that justice must be seen to be done prevailed, even if there is no evidence of any actual prejudice to litigants. Although he had been a dissenter in the Kress decision, the President of the European Court of Human Rights, Wildhaber, joined the majority in Martinie, refusing to overturn its previous case law.

The reaction of the Conseil is shown by an interview given by its then chief, Vice President Genevois.⁵⁶ He did not hide his view that the Strasbourg court made the wrong decision. The minority of the court in Martinie explicitly made the argument ‘if it ain’t broke, don’t fix it’ to argue that a practice that had secured justice for more than 170 years should not be overturned simply because it might be misunderstood or because it did not fit a ‘purist’ conception of a fair procedure.⁵⁷ All the same, the approach of the Conseil d’Etat was been described aptly as ‘partly submission and partly interpretative resistance’.⁵⁸ In part it changed its procedure, as did the Cour de cassation, to encourage greater communication of the commissaire’s arguments to the lawyers before the hearing; it also accepted changes in the rules of procedure that allow a party to object to the presence of the commissaire at the deliberation among the judges. But the resistance has come in the interpretation of ‘participation’. Contrary to the view of the majority in Martinie, the Conseil does not consider ‘presence’ synonymous with ‘participation’. As will be seen in Chapter 4, although the rapporteur public (as he is now called) in the tribunal administratif or in the Cour administrative d’appel does not retire with the judges, he is entitled to be present in the Conseil d’Etat unless the parties object (and their lawyers never do!). The final element in this saga was the relabelling of the commissaire du gouvernement as ‘rapporteur public’ in 2009 and a number of changes in procedure, allowing the parties to respond orally to the arguments of the rapporteur public. Increasingly, the rapporteur public mirrored the Advocate General in the Court of Justice of the European Union. Distinctive traditional French conceptions of fair procedure have had to change to meet contemporary European conceptions of what a fair procedure now demands. Whereas for Hamson in 1954 ‘this autonomy [of the Conseil d’Etat is] self-evident’, conceptions of transparency have moved on.⁵⁹ As an occasional additional judge of the European Court of Human Rights, Pacteau is sensitive to how the French system looks from the outside. He noted in 2009 that:

It is true that it seems bizarre to see the Government, largely master of the composition of the Conseil d’Etat, not to mention that its president is the Prime Minister (albeit as a matter of protocol, but all the same…). Indeed notably, because it is highlighted, many administrative judges pass through mixed and subtle careers. Not least, there is the duality of its functions, with that other original feature, at least to external eyes, that litigation on decrees or ministerial decisions made after consultation with it is reserved to [the Conseil d’Etat] …, remembering the ambiguous care with which the Conseil d’Etat ensures that consulting it has been genuine, effective and without abuse, over and above the normal requirements for consultations.⁶⁰

Commentators have perceived a much deeper change resulting from the European Convention. Madiot argued in 1991 that French society was ‘dominated by an administrative law subordinated to a mythical and indefinable public interest which only operates for the almost exclusive interest of the administration and which, too often, reduces the individual to the level of a subject’.⁶¹ Braconnier argues that the Convention’s emphasis on the individual, his claims against the state, and the subordination of the state and its discretionary power to the law undermines the authoritarian aspects of the French public law tradition.⁶² For the French, the Convention also does not respect the public law/private law distinction, which is central to their conception of administrative law – for example, in relation to the application of art. 6(1) on a fair judicial process, or on principles of liability. In the case of the latter, the Strasbourg court takes the view that interference with individual rights requires a minimum standard of protection whether the interference results from the act of an individual or of a public body.⁶³ This clashes with the French tradition of seeing public law issues as conceptually distinct because the reconciliation of the interests of the public and an individual is not the same as the reconciliation of two competing individual interests. With some exaggeration, Braconnier argues that the focus on the protection of individual rights constitutes ‘a legal earthquake’ which requires the French to reassess both their conceptual structures and values and their organisation in public law.⁶⁴ Lasser also viewed the debate about the commissaire du gouvernement as a challenge to the traditional French conception of public law.⁶⁵ The Republican tradition focused on the public interest as determined by the general will of the people expressed through the legislature and the executive. Administrative law structured and facilitated this. By contrast, the European Convention represents, in his view, a more ‘liberal’ model of competing individual interests and entitlements which the law has to regulate. Administrative law is no longer naturally aligned with the state, albeit as a moderating and supervising influence. It is much more a neutral arbiter where the state has no particular special position.

Braconnier also argued that the Convention enriches administrative law by providing a new source of general principles of law and this leads to a decline in importance of the administrative judge in protecting rights. As will be seen in Chapter 3, the Conseil d’Etat had a very distinguished role in developing ‘general principles of law’ as the foundation for the protection of fundamental rights, especially in the 1950s. But the Conseil constitutionnel (set up in 1958) and the European Court of Human Rights (since direct petition was allowed by France in 1981) have become major judicial forces in defining standards for the protection of human rights in France. This has inevitably reduced the role of the Conseil d’Etat, which is effectively (though not formally) a hierarchically inferior court. French administrative law has to look to constitutional law and European laws for authoritative statements. Although it can still act innovatively in declaring new principles, but it is no longer the principal driving force.⁶⁶

1.7 Reform of the Administration

France may be distinctive in the organisation of its administration and in its administrative law, but it is not unusual among developed countries. Particularly since the Second World War, France has been subject to what Christopher Hood called ‘megatrends’ in government.⁶⁷ In the immediate post-war period, France centralised rebuilding its economy and society through the Plan, but in the 1980s, this gave way to reliance on the free market. Nationalisation of key public services and, in the early 1980s, of the ‘heights of the economy’ spawned a large number of public enterprises and publicly owned private law enterprises. But from the Chirac government of 1986, privatisation became a major way of organising public services, not least under the influence of European Union law which wished to avoid some Member States closing off sectors of their economies to competition coming from other Member States. A further trend was the introduction of new public management. Hood⁶⁸ points to changes in the technologies of government and of delivering public services, changes in social expectations and changes in the operation of political parties as some of the reasons for changes in government which occurred across a range of developed countries. The Organisation for Economic Co-operation and Development (OECD) has been a major location for the exchange of ideas among the administrations of many developed countries. Its location in Paris has been helpful for providing the French administration, among others with intelligence on what has worked in other countries and what are the best ideas. It has monitored developments in a large number of countries on themes such as ‘the modernising of the state’, a label and a theme which has dominated French government discourse for the past forty years.⁶⁹ The themes have involved control and reduction of budgets, accountability for results and a different kind of face for public administration towards the citizen. These trends in public administration also fitted into the changes in social expectations following the protests of May 1968. The demand for a more democratic, responsive and accountable government gave particularly French impetus to the general trend to a more consumer-like relationship between the users and providers of public services. ‘History, culture and the level of development give different characteristics and priorities to governments.’⁷⁰

Through the modernisation programme, the face of the administration has changed increasingly because of technology. The rather authoritarian, bureaucratic and anonymous face has given way to trends of more open government. Technology enabled greater accessibility of the administration to the public. Information could be provided by the administration to the public and the public could interact more easily with the administration. France was idiosyncratic in developing Minitel for this purpose, before migrating to the Internet. This was combined with a more personalised interaction. There has long been a requirement that the administrator making a decision should sign the document. So it was relatively straightforward to ensure that the citizen knew the name of the administrator dealing with their case.

Transparency was an early requirement of the modernising French state. The Law of 17 July 1978 gave any person the right of access on demand to files, reports, minutes, statistics, decrees and circulars held by the administration. It also gave individuals the right to request documents concerning them individually. These rights were enforced by the creation of a commission for access to public documents (the CADA; see Chapter 2, Section 4.1). The Law of 11 July 1979 introduced obligations for the administration to give reasons for its decisions affecting individuals unfavourably. In particular, reasons have to be given where civil liberties are restricted, penalties imposed, conditions are imposed on an authorisation, existing rights are restricted or withdrawn, time limits set or benefits refused when the requisite conditions are met. This reversed the normal expectation in administrative law that there was no requirement to give reasons without a specific text requiring this.

Accountability was among the earliest issues in France after 1968. Changes in complaints were introduced in 1973 when the Médiateur, the French ombudsman (now called the Défenseur des droits) was introduced. Whereas the administrative courts were concerned with the legality of administrative decisions, the Médiateur added a further level of accountability in terms of failure to fulfil its mission or unfairness in the results achieved. (We will return to this office, which became a constitutional office in 2008, in more depth in Chapter 2, Section 5.) The extension of scope and the lack of charge for this service provided an independent check on the administration, alongside the judicial controls.

Holistic approaches to the procedures of public administration have been slow to emerge. Unlike the United States and Germany, which enacted comprehensive legislation on administrative procedure in 1945 and 1976, France did not have a comprehensive text on non-litigation administrative procedure until the Code on the Relations between the Public and the Administration (CRPA) was enacted in 2015. Even then, this is a compilation of texts, rather than a systematic framework for these relations. The first stage was the decree of 28 November 1983 on the relations between ‘the administration and its users’. This clarified the status of documents such as governmental circulars (on which users were entitled to rely) and the duty of the administration to withdraw unlawful decisions (without waiting for a court order). It set out a number of rules on administrative procedure. In particular, it required the administration to acknowledge receipt of correspondence, to identify the civil servant responsible for the file and to initiate the transfer of an incorrectly addressed request to the right administration. It also set out a right for a user to make observations before a decision was made. It also set out procedures for consultative bodies, thus enabling the public to participate more effectively in decision-making. The next stage was the circular of Prime Minister Rocard of 23 February 1989 on the ‘renewal of the public service’. It sought to empower civil servants by imposing fewer controls and allowing them more initiative, having better dialogue between staff unions and the administration, and encouraging a more welcoming culture towards those using a public service. In its fourth section, the circular tried to foster a service culture within the public service. Here technology, personalisation and treating users of a service as partners, rather than the objects of administration were seen as keys. The mood of approach continued in the following decade, leading to the Law of 12 April 2000 on ‘the rights of citizens in their relations with the administration’. There is already a change of terminology – the talk of ‘rights’, the use of the word ‘citizen’, rather than ‘the administered’ and the reversal of the order of the citizen and the administration. The vision was that the citizen was not just the passive recipient of benevolent administrative largesse, but an active participant in shaping an appropriate public service. In terms of content, in particular, it re-enacted the provisions of the decree of 1983, and improved the role of the Médiateur.

Initiatives on the ‘modernisation’ of the public service have been found in all governments since the 1990s. The CRPA of 2015 came as the culmination of attempts to make public administration adapt to the requirements of a changing society. In trying to change the culture, there was a deliberate attempt to reduce grounds of complaint against the administration. Although this work concentrates on the methods of judicial redress against administrative action, that is inevitably only a pathological view of administrative law. The success of administrative law is that it provides a framework of procedures, rules and authority which enables the public to be served in an appropriate way by the administration. That relies on the success of the non-litigation parts of the system which form a background to the system of administrative litigation. More of this will be seen in the discussion on standards of good administration in Chapter 7.

1.8 A Note about Case Citation

This book, like most French administrative law textbooks and articles, refers to parties by name. So we talk about leading cases such as Blanco or Nicolo. This remains the tradition of printed French case reports. Alas, French courts have taken the decision that the names of parties are personal data which are protected from online dissemination without their consent. As a result, the online Legifrance website and other similar official websites have anonymised the names of parties in cases, even in old cases. In addition, art. 33 of the Law of 23 March 2019 prohibits the reuse of data on the identity of judges and registrars in cases. So it is not possible to undertake the classic English approach of analysing the decisions or opinions of different judges. All the online reader will get is the number of the case. Accordingly, this book gives the official numbers of every case. French lawyers have as much trouble with this as do British readers. British jurists are unlikely to remember ‘House of Lords, 5 May 1932’ instead of ‘Donoghue v Stevenson’!

All the same, the availability of French cases free and online is a great benefit. There are two collections. Legifrance provides access to legislation and judicial decisions of all French courts. The Conseil d’Etat’s ArianeWeb is accessible from the Conseil d’Etat website and it provides more precise searching of administrative law decisions, conclusions of the rapporteur public, as well as to consultative opinions. This website provides links to the archive of Conseil d’Etat decisions back to 1821.