2. The constitutional doctrines of European law

The European legal order has come to create a much more binding and intrusive legal system and a more effective cross-border trading environment than the more primitive and often unreliable versions provided by many treaty systems. The obligations of ordinary international treaties, particularly trade agreements, are commonly enforced by the practice of inter-state reciprocity and the use of ‘self-help’ in the form of ‘tit-for-tat’ retaliation between trading partners.Footnote ¹¹ However the obligations of the Treaty of Rome, as it has come to be interpreted, are enforced instead by the domestic courts of the European states.Footnote ¹² According to the doctrine of direct effect, sufficiently clear and precise European obligations can be enforced by private parties before the domestic courts of their own states. According to European law’s supremacy doctrine, domestic courts must apply directly effective European law in place of conflicting national law. Because of European law’s separation from ordinary international law, any recourse to inter-state reciprocity or ‘tit-for-tat’ retaliatory actions between the European states is completely ruled out. These are the essential doctrines of European law, often referred to as providing a ‘constitution’ for Europe, reflecting the ways that constitutional laws ‘trump’ or displace the application of ordinary law in many political systems.Footnote ¹³

In more detail, a person or business may be adversely affected by a member state’s non-fulfilment of a European obligation – a treaty provision, or a European rule negotiated by the governments of the European states within the European institutions. (These European obligations originally overwhelmingly related to the construction of the intra-European market and customs union, however over time they have come to cover an ever wider range of non-economic issues, including social and environmental regulations as well as police cooperation.) The person or business can bring a case before a national court demanding that their national judge directly enforce the European law obligation. The national judge may send a so-called ‘preliminary reference’ to the Court of Justice in Luxembourg, seeking the Court’s interpretation of the disputed European law provision. According to the doctrines of direct effect and supremacy, the national judge then directly applies European law in their own court, in line with the ECJ’s interpretation, regardless of whether there is a domestic legal provision providing for that outcome, and indeed regardless of whether the European law obligation may conflict with national statutes or constitutional provisions. Ordinary individuals and businesses do not have the right to demand the enforcement of the UN Charter, or the rules of the WTO, through their domestic courts working in organised cooperation with a treaty-based tribunal. By contrast, individuals and businesses in the EU – in Germany, France, and all the other member states – can, and do, use courts in their own states, working with the ECJ, to obtain the domestic enforcement of EU law.

These doctrines, together with a variety of less strictly doctrinal factors, such as the growing availability over time of judges and lawyers with the competence and willingness to make use of these European litigation opportunities, built the European legal order. In the early years, the observance of European law by the national courts was limited indeed and the Court of Justice in Luxembourg only heard only a handful of disputes from the original six member states (France, West Germany, Italy, Belgium, Netherlands, and Luxembourg). These days the Court hears several hundred cases annually, and stands at the peak of many thousands of European law judgements made on a wide range of topics by national courts and tribunals across 27 states, with over 400 million inhabitants.

The constitutional doctrines of European law have always been controversial. The fundamental objection is a democratic one: citizens in the European states should be subject to the laws made (and re-made) by their own national parliaments, their primary forum of involvement in politics and choices over self-government. Indeed, treaty obligations derived from more ordinary treaties are usually only legally applicable in the domestic orders of participating states to the degree that their national parliaments have so provided in domestic law. These parliamentary rights are purposively disrupted by the direct effect and supremacy doctrines of European law which claim that European law imposes obligations ‘independently of the legislation of the member states’, in the famous words of the ECJ’s 1963 Van Gend en Loos judgment. As Weiler explained in a celebrated Yale Law Journal article:

Of course, the constitutional doctrines of European law bind all political institutions in the member states, including national executives. There is a difference however, in that national executives participate directly in the making of European rules through the dominant EU decision-making body, the Council of Ministers, whereas national parliaments have at best an indirect and limited role in trying to control European negotiations by their national governments, often with incomplete information, and cannot prevent the national application of directly effective European rules that national executives have agreed within the European institutions.Footnote ¹⁵ As Vaubel explained, ‘the European Union enables the member governments to evade parliamentary control … Whoever used to believe that democratic governments are controlled by Parliament must recognise that, in the European Union, the member-governments, assembled in Council, control their parliaments’, or, in Weiler’s formulation, the governments of the member states ‘often act together as a binding legislator outside the decisive control of any parliamentary chamber’.Footnote ¹⁶ These legal doctrines have therefore directly lead to the increase in executive power, and the decrease in national parliamentary control, that is, for many observers, the ‘first and foremost’ basis for claiming that the European Union suffers from a ‘democratic deficit’.Footnote ¹⁷

The constitutional doctrines of European law have their defenders, of course. As has been repeatedly observed, given the attenuated form of state consent to EU activities, the role of the European institutions is most commonly justified not by ‘input’ criteria such as democratic control through elections featuring competing political parties, but rather by ‘output’ criteria, such as ‘effectiveness’.Footnote ¹⁸ That is definitely true in the case of the European legal order, which primarily justified by pointing to the defects of the obvious alternative – the ‘archaic’ and unreliable system of classical international law as well as reciprocity-based forms of international trade law, which effectively facilitate states in evasions of legal obligations, unilateral opt-outs and non-compliance.Footnote ¹⁹ Against this background, the distinctive focus of Europe’s ‘transnational constitution’ is therefore neither the organisation of popular sovereignty nor the protection of marginalised minorities, as theorists normally understand the function of constitutions, but improving governmental effectiveness in matters of common European interest.Footnote ²⁰

Where then did the constitutional doctrines of European law come from?

3. The origins of the constitutional doctrines of European law

One might assume that the major doctrines of European law, the ones that distinguish it most clearly from other international treaty systems, would have been explicitly set out in the Treaty of Rome that established the European Economic Community in 1958. This is not the case, however, as the Treaty did not provide for the direct effect of European obligations, for their supremacy in the national legal orders, or prohibited inter-state retaliation between the European states. To be sure, the Treaty did provide for the Court of Justice itself (Articles 4 and 164), for the preliminary reference procedure by which national courts could submit questions to the Court of Justice (Article 177), for the direct application in the domestic legal orders of Regulations made by the European institutions (Article 189), and that any safeguard measures (in effect, derogations from the Treaty) adopted by the member states were required to be authorised, in advance, by the independent European Commission (Article 226).Footnote ²¹ All these provisions pointed towards an ambitious institutional framework, including limited elements of direct application of European obligations in the domestic legal orders, but they did not set out the constitutional doctrines of European law.

The European Court of Justice itself declared these doctrines in a series of major judgements in 1963 and 1964, often against the opposition of lawyers representing the member states.Footnote ²² In Van Gend en Loos in February 1963, the ECJ declared that private parties could enforce European obligations before national courts in their own member states.Footnote ²³ In Costa v. ENEL in July 1964, the ECJ declared that European law must be applied by national courts in the place of conflicting national legislation.Footnote ²⁴ In Dairy Products in November 1964, the ECJ declared that member states in the European Economic Community were prohibited from making any use of the ‘self-help’ retaliation and reciprocity behaviours which are so central to the functioning of more ordinary forms of international law.Footnote ²⁵ Through these three decisions, and the many others that confirmed and expanded on them throughout the 1960s and 1970s, the ECJ was itself the author of the constitutional doctrines of European law. To be sure, in each case, the Court was responding to the cases that came before it, and often to the urging of the Legal Service of the European Commission, and a variety of pro-European litigants, to make ambitious judgements.Footnote ²⁶ It was however the Court itself that declared the major doctrines that laid the foundations of today’s mighty European legal order.

It is therefore critical to understand the reasons behind the judges’ decisions. Attempts to peel the onion one stage further, however, run into a formidable set of obstacles. Unlike the United States Supreme Court, the ECJ publishes no dissenting or concurring opinions to supplement the Court’s aggregate judgment always signed by all judges. ECJ judges are sworn to secrecy and the Court’s archives recording its internal deliberations remain largely closed, even in relation to judgements made more than fifty years ago. The Court of Justice does not facilitate the study of the contributions of individual judges to its internal decision-making.

Recently, however, there has been a little progress. Aspects of the Court’s internal debate in the 1963 Van Gend en Loos case were first revealed by former judge’s assistant Paolo Gori.Footnote ²⁷ Gori’s documentation demonstrates that Andreas Donner, Dutch judge and from 1958 to 1964 President of the Court of Justice, opposed the declaration of the direct effect doctrine. Donner’s opposition was however overcome by Italian judge Alberto Trabucchi and French judge Robert Lecourt, both new appointees, who mustered a majority of judges’ votes in favour of the direct effect doctrine and the Van Gend en Loos judgement as history now knows it (Donner himself resigned as President of the Court the day after the 1964 supremacy judgment, noting that he had lost the support of his colleagues).Footnote ²⁸ Gori’s revelations, by disclosing the priority influence of some judges more than others, enabled a much greater engagement with the intersection of judicial biography and the Court’s output.

It is therefore to Trabucchi and Lecourt, rather than to Donner, that the primary influence on the Court of Justice’s major judgements in 1963 and 1964 must be attributed. And it is Lecourt in particular who became the dominant influence on the Court for many years, eventually serving as President of the Court from 1967 to 1976. Lecourt is indeed the only ECJ judge (apart from Donner) who served throughout the crucial 1962 to 1976 period. Lecourt’s contemporary writings outlined important doctrinal developments well in advance of the Court’s actual judgements,Footnote ²⁹ and publicly advocated for the Court’s role as a ‘motor’ of European integration as early as 1963.Footnote ³⁰ Another prominent judge of the European Court, Pescatore, talked of the ‘jurisprudential miracle’ of the Court’s ‘Lecourt years’ [années Lecourt], when so many of the Court’s ‘revolutionary’ decisions were made.Footnote ³¹ It is Lecourt who travelled Europe tirelessly propagandising the ECJ’s new doctrines and it is Lecourt’s 1976 book, L’Europe des Juges, alone of the publications of the Van Gend en Loos judges, that can still be found on the shelves of European law professors across the continent.Footnote ³² The Court of Justice in the 1960s and 1970s has come to be recognised as the ‘Lecourt Court’ in much the same way that legal historians in the United States might refer to the ‘Warren Court’.Footnote ³³ Lecourt appears to have occupied, for the ECJ, the role of the dominant judge during the period of its most fundamental judgements. The same role that was played by Chief Justice John Marshall on the early United States Supreme Court.Footnote ³⁴ As Gragl writes, ‘all decisions of the Union’s foundational period can be attributed … to the arrival of Robert Lecourt’ at the ECJ one year before the Van Gend en Loos judgement.Footnote ³⁵

To be sure, the mere declaration of these doctrines by the ECJ does not explain their eventual success within the law and politics of the European member states. Early responses to this question revolved around competing ‘big picture’ sketches that supplemented discussion of legal texts and an anonymous court with various aspects of the wider context, either the incentives of ‘lower’ national courts to empower themselves by applying European law domestically, or the incentives of the European states to obtain a treaty enforcement institution that would be responsive to the demands of the major member states but which states could nonetheless ‘ignore’ when necessary.Footnote ³⁶ Recent historical and political science research however has made great progress on more specific and empirically-grounded answers to this question. On the strictly legal side, national courts (frequently appeals or other ‘higher’ courts, in fact, but not national constitutional courts) proved increasingly willing over time to directly apply European law and submit preliminary references to the Court of Justice, often initially at the instigation of a small number of ideologically committed Euro-lawyers.Footnote ³⁷ The Legal Service of the European Commission, under its director Michel Gaudet, propagandised in favour of the new European law doctrines by networking with national judges and private sector Euro-lawyers.Footnote ³⁸ When judgements of the Italian and German Constitutional Courts in 1965 and 1967 raised the possibility that European law would not be applied domestically because of fundamental rights concerns, the ECJ moved to defuse the issue by itself taking on a role reviewing European law on fundamental rights grounds.Footnote ³⁹

On the political side, the famous judgements of 1963 and 1964 had in fact been supported by senior German policy-makers, despite contrary arguments advanced by Germany’s lawyers before the Court itself.Footnote ⁴⁰ The wider political and economic conditions of post-war Europe were also conducive to an intrusive and binding dispute settlement system, much harder for states to ignore than ordinary international law, because high levels of economic interdependence combined with generous welfare states and intra-industry trade that facilitated the necessary adjustments by adversely affected industries and groups.Footnote ⁴¹ The fact that governments were firmly in control of the ECJ’s judicial appointments, but did not attempt to use them to influence the Court, suggests that the European governments were willing to accept unprecedented constraints on state policy.Footnote ⁴² Wider trends in European political development favouring administrative delegation and decree-powers domestically also facilitated the international delegation of powers to the European institutions.Footnote ⁴³ Newer political science research on the ECJ, frequently quantitative, often focusses on contemporary issues, such as the Court’s responsiveness to the legal arguments of the member states, changing regulatory and litigation practices, and the impact of the latest (post-2008) judicial appointments procedure.Footnote ⁴⁴

The focus of this paper however is on understanding the constitutional thinking of the ECJ’s leading judge during Europe’s legal revolution in the 1960s and 1970s. We know the output of the Court of Justice in its ‘Lecourt years’, but we know little of Lecourt himself. Why did he support and promote the Court’s famous judgements? Here again the obstacles can appear forbidding. Lecourt wrote no detailed account of his time at the Court and his private papers were apparently destroyed at his own insistence prior to his death.Footnote ⁴⁵ As a judge at the birth of a new legal order, Lecourt’s situation precludes examining earlier service in an appeals court (or as a prominent litigator) to inform our understanding of later judicial behaviour after a supreme court appointment. Lecourt had not even been a judge in France. As Morten Rasmussen writes, ‘Lecourt’s legal thinking is difficult to trace, because as a politician, he did not publish regularly on legal matters’.Footnote ⁴⁶ Lecourt is often described as a Christian Democrat, and also as a ‘conviction European’,Footnote ⁴⁷ because of his long association with ‘pro-European’ activism in France and across Europe.Footnote ⁴⁸ An emphasis on the latter factor fails however to sufficiently distinguish Lecourt from say Donner whose speeches as President of the Court were often themselves distinctly ‘pro-European’, but who nonetheless failed to support the doctrine of direct effect when it came to the Court in 1962.Footnote ⁴⁹ Moreover, it is only in hindsight, given the way that European law developed, that ‘Europe’ came to be understood as so clearly bound up with these particular legal doctrines. Rather than looking to Lecourt’s support for ‘Europe’ in the abstract, what is needed is evidence for why Lecourt would have supported, more specifically, the doctrines of direct effect and supremacy, as well as European law’s rejection of inter-state retaliation.

There is in fact evidence that Lecourt was personally committed to the rejection of inter-state retaliation prior to his appointment to the Court. Lecourt’s doctoral dissertation in law, published in 1931, was a persistent and unswerving critique of the use of ‘self-help’ in French property disputes, which then turns in its final pages into a full-throated call for international organisations to likewise prohibit the use of ‘self-help’ in international politics, in wording similar to that later used in the ECJ’s 1964 Dairy Products judgement.Footnote ⁵⁰ This is an issue-area where a clear continuity can be traced between a judicial appointee’s earlier legal commitments and the doctrines that a court advanced after their appointment.Footnote ⁵¹

What about Lecourt’s support for European law’s direct effect and supremacy doctrines? At least three aspects can be identified which would incline against, or in favour of, these new doctrines.

First, as scholarship rightly emphasises, the direct effect and supremacy doctrines were ‘revolutionary’ declarations by the Court, involving a decisive break with long-standing approaches to international law and treaty politics.Footnote ⁵² Common forms of legal conservativism – in the sense of attachment to and boundedness by inherited procedures and institutions – would caution against such a leap into the unknown.

Second, the direct effect and supremacy doctrines of European law had the obvious potential to displace ordinary forms of law-making in the national parliaments of the member states. Weiler’s account of the direct effect and supremacy doctrines, above, describes them straightforwardly as a mechanisms for removing the ability of national parliaments to vote legislation. A common attachment to the authority of national legislatures to vote on the laws affecting their citizens’ lives would therefore have inclined against the direct effect and supremacy doctrines.

Third, and on the other hand, the value that might encourage support for the new doctrines of European law would be an uncommon emphasis on ‘effectiveness’. ‘Output effectiveness’ is, after all, the central justification for activities of the European institutions. A strong predisposition to the argument that public institutional structures must first and foremost provide workable solutions to the practical challenges facing them, with other values sidelined if necessary, would likely incline towards the support of these doctrines.

It is no surprise that many have balked at the direct effect and supremacy of European law. The puzzle is rather, why did judge Lecourt support them? What was there in his own background to make him open to accepting, developing, and promoting the ambitious, ‘constitutionalising’ doctrines put forward by the Legal Service of the European Commission and a variety of pro-European private litigants? What, in short, was Lecourt’s constitutional ideology? To answer this question, we turn to the turbulent politics of postwar France, where, half-hidden and half-forgotten, Lecourt as a politician had already played a key role in designing constitutional structures.

4. Article 49-3 of the French constitution

France’s current constitution – the ‘Fifth Republic’ – was put to the French people for approval by referendum in September 1958 by the government of Charles de Gaulle, coming into force on 4 October 1958. There were many changes compared with the previous constitution of Fourth Republic France (1946–1958), however some were more important than others. Professor Georges Vedel liked to say that the Fifth Republic consisted of three essential elements : the direct election of the President by universal suffrage, the creation of the Constitutional Council, and Article 49-3.Footnote ⁵³ The direct election of the President by universal suffrage – not actually provided for in the 1958 Constitution but approved by a subsequent referendum in 1962 – gave the President a democratic mandate separate from, and perhaps superior to, the National Assembly. The Constitutional Council was in effect a constitutional court, a novelty in republican France which had long refused to allow judicial control of legislation.Footnote ⁵⁴ It is however the last of Vedel’s three essential elements of the Fifth Republic – Article 49-3 – to which we will now turn.

Article 49-3 of the French Constitution as enacted in 1958 stated as follows:

At first sight, this provision might look like the bland constitutionalisation of a ‘confidence vote’ procedure, common in many parliamentary systems, whereby the government seeks a confidence vote on an item of parliamentary legislation, with the implied threat of government resignation and/or new elections if the government loses the vote. The confidence vote organised by the British Conservative government on 23 July 1993 to ensure the ratification of the Maastricht treaty would be one prominent, and Europe-related, example. Article 49-3 however contains distinguishing features which can only be understood by setting out some French political history.

Under the pre-1940 ‘Third Republic’ constitutional system, the government could be removed from office by losing a simple, or relative, majority vote in the Chamber of Deputies. Aiming to reinforce the stability of governments, the new, post-war, 1946 Constitution provided that the government could only be overthrown when an absolute majority of deputies had opposed the government. Repeated overthrows of governments by absolute majorities could then prompt a dissolution and new elections. This arrangement however opened up a particular scenario whereby the government called a vote of confidence on an item of legislation, and the parliamentary vote by the deputies actually voting went against the government – but not by the required absolute majority of deputies that would constitutionally oblige the government to resign.Footnote ⁵⁵ As a result, the government was constitutionally entitled to remain in office, however, the proposed legislation on which the government had staked its responsibility was not adopted into law. So the government appeared powerless and often felt politically obliged to resign despite not being constitutionally obliged to do so, and without the constitutional requirement for a dissolution and new elections being reached. Many of the ministerial crises of the Fourth Republic system were the result of this sort of parliamentary manoeuvring, whereby deputies belonging to government parties would abstain in confidence votes on difficult issues, allowing the government to ‘survive’ (at least according to the constitutional rules) but without controversial legislation being adopted.Footnote ⁵⁶ This scenario was widely understood to be a central cause of the persistent ministerial instability in the Fourth Republic, which had indeed 21 governments over its short 12 year life.

Article 49-3 was designed to address this difficulty. Under this procedure, the government engages its responsibility on a legislative text, after which it is for the opposition to respond and a variety of scenarios may occur. If no vote of censure is tabled by opposition deputies, the text is considered adopted without any vote at all. If a vote of censure is tabled, only the votes in favour of the censure are tallied, so that abstention favours the government and therefore can no longer be meaningfully carried out. If the votes in favour of censure reach the necessary absolute majority against the government, then the government must resign and the text, of course, is not adopted. If the votes in favour of censure fail to reach an absolute majority against the government, then the government survives and the text is at the same time considered adopted. There is no longer any scenario whereby the government constitutionally survives but the text on which the government has engaged its responsibility is not adopted – the government and the text now stand or fall together. As commentary on the new 1958 constitution was quick to point out, the logic of Article 49-3 was that rejection of the government’s proposed legislation must be explicit, while acceptance need only be implicit.Footnote ⁵⁷

The vote of confidence procedure set out in Article 49-3 continued to be relevant after 1958, despite the increase in the independent constitutional powers of the President, because under the Fifth Republic the French government (and its Prime Minister) can still be forced to offer their resignation by an absolute majority of deputies, and continues to need the approval of the National Assembly to pass laws in many important issue areas. So just as under the Fourth Republic, Fifth Republic governments need a mechanism for ensuring that legislation can be passed. As Michel Debré, principal drafter of the new constitution, explained in his elaboration of the text before the Conseil d’Etat: ‘Experience has led, moreover, to provide a somewhat unusual disposition to ensure, despite manoeuvres, passage of an essential piece of legislation’.Footnote ⁵⁸

The purpose of this ‘somewhat unusual disposition’ – Article 49-3 – was to prevent ministerial crises resulting from the inability of the French government to pass legislation that it considers indispensable to its continuance in office. On this score Article 49-3 has been a remarkable success, as the short-lived and crisis-ridden governments characteristic of the Fourth Republic, with 21 governments over 12 years, from Ramadier to De Gaulle – have been replaced by the greater stability of governments characteristic of the Fifth Republic, with 26 governments over 65 years, from Debré to Attal. To be sure, other institutional changes have contributed to this greater governmental stability, above all the election of the President by universal suffrage, whose party often goes on to considerable success at subsequent legislative elections. However the French parliament – like French society – remains divided and fractious, with many ideologically conflicting political parties whose deputies in turn may demonstrate less than fully reliable obedience to party leaders.Footnote ⁵⁹ In these circumstances, Article 49-3 has been used 113 times since 1958 to successfully ensure the adoption of legislation by the National Assembly, with usage varying of course according to political conditions.Footnote ⁶⁰

Article 49-3 has from the first been a controversial aspect of the 1958 Constitution. As noted above, it is unusual among the various confidence vote processes in parliamentary systems, in permitting a law to be adopted without a normal parliamentary vote, or in the context of a vote where a majority of recorded votes are against the government and its legislative text. In the United Kingdom’s 1993 Maastricht vote, for example, the Conservative government needed to ensure that its own party MPs voted in favour of its confidence motion approving the Government’s policy on the Treaty’s ‘Social Chapter’, and was successful with 339 votes in favour and 299 against. However Article 49-3 purposefully makes no such requirement. In the case of the legislation initiating the French nuclear programme (the ‘force de frappe’) in 1960, where Debré’s government had engaged the responsibility of the government in line with Article 49-3, Le Monde reports the opposition vote of censure required a majority of 277, and obtained only 207 votes, with the result that the law was considered adopted.Footnote ⁶¹ There were no recorded votes ‘in favour’. Article 49-3 has regularly been called upon by governments to obtain the adoption of legislation in sensitive areas such as social security and the age of retirement, including recently in 2023 by a government appointed by President Macron, prompting hundreds of thousands of protestors to take to the streets, sometimes violently.Footnote ⁶²

Paul Reynaud, former Prime Minister and a persistent detractor, described Article 49-3 as ‘degrading to the National Assembly’.Footnote ⁶³ It is widely characterised as ‘undemocratic’ and contrary to the principle that the National Assembly, chosen by the people, must vote to approve new laws.Footnote ⁶⁴ Socialist politician François Hollande called it ‘a brutality, a denial of democracy’Footnote ⁶⁵ , and critics describe it as a means of ‘disciplining the parliamentary majority without trying to persuade it’.Footnote ⁶⁶ Article 49-3 is controversial outside the political elite as well, with 2016 polls highly critical of its use for contentious employment legislation.Footnote ⁶⁷ Is there any article of the 1958 constitution which is as roundly criticised as Article 49-3? asks Aromatario.Footnote ⁶⁸

Article 49-3 has its ardent defenders, however. These do not tend to vindicate the supposed democratic credentials of Article 49-3, but argue instead that the practical results of getting rid of it would be disastrous. Public justifications of Article 49-3 emerged in force during a constitutional revision project in 2008, which limited the use of 49-3 to budgetary and social security matters, as well as (after much debate) one additional project per parliamentary session. ‘How did the constitutional reform committee come to have the absurd idea of practically annihilating the centrepiece [la pièce maîtresse] of our constitution?’ wrote Ameller.Footnote ⁶⁹ ‘What is at risk is nothing more or less than … the survival of the Fifth Republic’, wrote Jan, ‘Looked at correctly, the new provision contains the extraordinary potential to destroy the principal accomplishment of Michel Debré : governmental stability. … Without 49-3, the legislature would be a nightmare for the executive’.Footnote ⁷⁰

Given its profound impact, there has long been interest in the origins of this procedure. As one of the early polemics against the 1958 Constitution put it, ‘which eccentric came up with that idea?’ (quel est le farfelu qui a inventé [ça]?)Footnote ⁷¹ Where then did ‘Le 49-3’ come from?

5. The origins of Article 49-3

France’s current ‘Fifth Republic’ Constitution was adopted on 4 October 1958, replacing the ‘Fourth Republic’ Constitution that been adopted in 1946. It was initiated by De Gaulle who had returned to power in May 1958, initially as Prime Minister within the existing Fourth Republic framework, to address the crisis in Algeria and the threat of a military coup, only on condition that the political parties supported a constitutional revision to install a more Presidential form of government. De Gaulle is therefore popularly recognised as the creator of the 1958 Constitution. As the website of the French National Assembly announces, ‘the institutions of the Fifth Republic, set up in 1958, correspond to the ideas of General de Gaulle, as he had set them out in 1946’.Footnote ⁷² De Gaulle had indeed outlined his constitutional ideas in a speech in Bayeux on 16 June 1946,Footnote ⁷³ where he had advocated for a Head of State, chosen outside parliament by a wider constituency, with the power to name government ministers and act as an ‘arbiter’ in national politics. This would leave parliament to vote laws and the budget, without however the parliamentary domination of the executive that had, in De Gaulle’s view, contributed mightily to the weakness of the French state.

The lead drafter of the 1958 Constitution was Debré, a long-time associate of De Gaulle, who became Minister of Justice in De Gaulle’s government in June 1958. Debré had served as a Senator for many years, from which position he heatedly castigated the workings of the Fourth Republic, and advocated for a strong Prime Minister and a ‘rationalised parliament’ inspired in part by the British model.Footnote ⁷⁴ De Gaulle himself presided over the interministerial committee where draft provisions of the new Constitution were proposed, rejected, or agreed, and then Debré (with a few collaborators) produced a complete draft text, which after further consultation with a committee of parliamentarians and the Conseil D’Etat (but little substantive change) became the final text put to the voters on 28 September 1958.

Since Article 49-3 is such a conspicuous part of the Fifth Republic Constitution, not surprisingly it is often publicly associated with the persons of both De Gaulle and Debré. In debates over 2008 constitutional amendments that eventually came to revise the original 1958 wording of Article 49-3, commentators frequently associated possible limits on Article 49-3 as an attack on Debré’s major achievements, or as a rejection of De Gaulle’s Fifth Republic itself. Similarly, Prime Minister Borne declared during the March 2023 use of the 49-3 mechanism to obtain the approval of a raise in the French age of retirement, ‘The 49-3 mechanism is not the invention of a dictator, but the profoundly democratic choice made by General de Gaulle and approved by the people of France’.Footnote ⁷⁵

Article 49-3 was not initially advocated by either De Gaulle or Debré, however.Footnote ⁷⁶ There is no mention of any such mechanism in De Gaulle’s 1946 Bayeux speech, and René Cassin, another long-time ally of De Gaulle, told the Conseil d’Etat on 17 August 1958 that draft article 45 [later 49] ‘unfortunately’ had a ‘parliamentary’ origin and that de Gaulle had gone to ‘all the trouble in the world’ to keep it out of the constitution.Footnote ⁷⁷ As for Debré, he was opposed to Article 49-3, and attempted to have the mechanism removed from the agreed draft.Footnote ⁷⁸ It is from the initiative of other members of the interministerial committee that what became Article 49-3 was introduced into the draft constitution of the Fifth Republic.Footnote ⁷⁹

The interministerial committee was attended by powerful parliamentarians of the Fourth Republic regime, including Pierre Pflimlin (MRP – Christian Democrat) and Guy Mollet (Socialist), who now served as Ministers of State in De Gaulle’s constitution-writing government. Among historians of the 1958 Constitution it is widely accepted that it was at the insistence of these two former Prime Ministers of France, with rather more experience in executive–parliamentary management than Debré and De Gaulle, that Article 49-3 appears in the 1958 Constitution.Footnote ⁸⁰ In doing so, they drew on proposals to reform the Fourth Republic constitution that had been advanced in early 1958 by the administrations of Félix Gaillard and then again by Pflimlin himself (Gaillard’s successor), and which had been approved (as the first stage of the process) in a vote of the National Assembly in March 1958.Footnote ⁸¹ Article 49-3 is therefore neither especially ‘Gaullist’ nor particularly ‘Fifth Republican’ in its origins, as it was a constitutional reform already being considered separately from De Gaulle’s return to power and France’s turn to a more Presidential form of government.Footnote ⁸²

The origins of Article 49-3 therefore lie in the reforms proposed by the Gaillard government in early 1958. The primary author of these proposals was not Gaillard, however, but Robert Lecourt, leader of the MRP [Mouvement Républicain Populaire] Christian Democratic parliamentary group, and Minister of Justice in Gaillard’s government.Footnote ⁸³

It was Lecourt who had advocated similar constitutional changes, to allow legislative texts to be adopted where a confidence vote failed to achieve an absolute majority against the government, over the years prior to the formation of the Gaillard government.Footnote ⁸⁴ It was Lecourt who conducted the ‘round table’ consultations on constitutional reforms with party leaders. It was Lecourt who led the defence of the reform in the public arena, in the press, and on the radio. It was Lecourt who advocated the constitutional proposal in a series of tumultuous parliamentary debates throughout the early months of 1958. Lecourt’s leadership on the constitutional reform package (and particularly on the confidence vote procedure) was widely acknowledged at the time, with the plans referred to as ‘Gaillard–Lecourt reforms’ or even just the ‘Lecourt reforms’.Footnote ⁸⁵ Lecourt was the ‘theoretician’ [doctrinaire] of the reforms, according to the Gazette de Lausanne,Footnote ⁸⁶ and is described by scholar de Bujadoux as the ‘true animator’ [véritable inspirateur] of the government’s proposals.Footnote ⁸⁷ More than Pflimlin, Mollet, or Gaillard, and certainly more than either De Gaulle or Debré, Lecourt was the main actor behind the restructured relationship between the French government and the National Assembly that was eventually ratified into the 1958 Constitution.Footnote ⁸⁸ As La Croix’s correspondent on the constitutional process noted late in the summer of 1958, ‘The proposed system [for the vote of confidence] is, in fact, close to that advocated by M. Lecourt, when he tried to reform the previous Constitution, a system aimed at preventing governments from being overthrown by a minority, with abstentions counting against it. This will undoubtedly be one of the best provisions of the institutions of the Fifth Republic.’Footnote ⁸⁹

The parliamentary sessions of early 1958 debating Lecourt’s proposals had in fact many similarities with later public discussions of Article 49-3, with critics alleging that the proposed constitutional reforms were ‘undemocratic’ while their defenders, above all Lecourt, emphasised the need for constitutional structures to ensure effective government institutions.

Lecourt’s plenary speech began by insisting on the permanent instability caused by the then current constitutional rules:

Opponents of the reform by contrast claimed that the proposed reforms were undemocratic and overly emphasised an unprincipled form of effectiveness. Paul Reynaud gave an early rehearsal of his long-standing opposition to what would become Article 49-3 with his critique of Lecourt’s proposals:

Reynaud contrasted the proposal with German constitutional arrangements which did indeed put limitations on a purely ‘negative’ vote of confidence, however the German Bundestag continued to vote to enact legislation in the normal way. Agreeing with Reynaud, Raymond Triboulet, deputy and later Gaullist minister, said that the proposals were ‘at first sight incredible’, demonstrating a ‘certain contempt of parliamentarism … how could republicans possibly support them?’, while leftwinger Pierre Cot described the proposed reform as ‘shocking’ [choquantes].Footnote ⁹²

Paul Anxionnaz, a Radical party deputy, characterised the proposals as ‘pure pragmatism’, without regard to democratic theory:

These criticisms Lecourt deflected more than denied, justifying the proposals as not too different from other aspects of French law-making with limited direct parliamentary involvement, and emphasising again and again the overriding need for stability and effectiveness:

Lecourt’s speech ended with a flourish:

Lecourt’s proposal, in modified form, won the support of the National Assembly in March 1958, and, modified once again, became Article 49-3 of the 1958 Constitution. The differences between the various proposals were consequential, as they sought to prevent abstentions in a variety of ways, and also supplemented the ‘engagement of responsibility’/‘vote of censure’ mechanism with a variety of additional requirements, for example for the motion of censure to suggest an alternative prime minister.Footnote ⁹⁶ Nonetheless the essential mechanism contained in the 1958 Constitution (not these complicating additions) was always the central focus of the early 1958 proposals reforming the confidence vote system. Indeed in a radio interview on 12 January 1958 Lecourt defends exactly the mechanism as it later came to be set out in Article 49-3 : the government could engage the responsibility of the Government before the National Assembly on the vote of a text, and if no vote of censure was tabled, or if a vote of censure failed to be adopted by an absolute majority of deputies, the text would be considered adopted.Footnote ⁹⁷ That is the direct thread from Lecourt’s January 1958 proposals to the famous provision of the 1958 Constitution.

6. Effectiveness, parliamentary democracy, and European law

The European legal order is a world historical novelty, a momentous transformation and legalisation of international politics that began in postwar Western Europe and now includes 27 European states. The essential organising principles of this new legal system had not been negotiated into the Treaty of Rome by the European states however, but were declared instead by the European Court of Justice itself throughout the 1960s and 1970s, above all in the Court’s ‘revolutionary moment’ in 1963 and 1964. Previous research has come to recognise French politician Robert Lecourt as the leading judge on the Court throughout this extraordinary period. At that point however progress has stalled, given the closure of the Court’s archives on its internal decision-making, Lecourt’s failure to publish regularly on legal matters, and the apparent destruction of Lecourt’s own private papers. Thus the European legal order is frequently explained as the creation of an anonymous and personality-less ‘court’, even though it is widely acknowledged by lawyers, legal historians, and political scientists alike that courts are in fact controlled by judges, powerful individuals whose judicial decisions often demonstrate considerable continuities with their earlier political and constitutional commitments.

This paper has shown, first, that judge Lecourt’s prior political and constitutional values can indeed be identified, through his involvement in French politics. Far from a conventional French deputy, Lecourt was a specialist in institutional and constitutional design, open to radical changes in the way votes were counted, and in the allocation of agenda-setting rights and legislative powers, as required to achieve his desired results. More specifically, he was a veteran campaigner for the reduction or removal of the French parliament’s role in national law-making, to promote the stability and effectiveness of the French executive. While certainly a committed republican and democrat, as demonstrated by his background in the anti-Vichy resistance, Lecourt was moreover willing to go further than many contemporary critics of France’s parliamentary system. Consider that Debré, a determined critic of the French legislature if ever there was one, initially found that the proposals for Article 49-3 ‘offended his sense of parliamentarism’.Footnote ⁹⁸

So it turns out that the half-hidden, half-forgotten leading judge behind Europe’s legal revolution was also the half-hidden, half-forgotten leading instigator of perhaps the most controversial provision of France’s 1958 Constitution.

Second, for explanations of the birth of the European legal order, this paper has revealed continuities between Lecourt’s approach to reforming France’s constitution and the jurisprudence of the Court of Justice after his appointment. For Article 49-3 and the constitutional doctrines of European law share elements of a common ideology. Both were revolutionary novelties, with little background in French constitutional practice or classical international law. Both were rule-making procedures commonly justified by claims of ‘effectiveness’, as a pragmatic improvement on sub-optimal previous arrangements (the instability and paralysis of French governments, the weaknesses of ordinary international law).Footnote ⁹⁹ Both were also frontal assaults on widely prevailing understandings of French (and, in the European case, other nations’) democratic procedures, that law could only be made by majority votes of parliamentarians to approve texts in the ordinary way. Indeed, recalling Paul Reynaud’s claim that if Lecourt’s confidence vote procedure was accepted, the French National Assembly would be the only parliament that would not vote to enact the law, it is equally true that, as a result of the constitutional doctrines of Lecourt’s Court of Justice, the other national parliaments in Europe no longer vote to enact the law either.

The broad continuities between Article 49-3 and the constitutional doctrines of European law appear to have been acknowledged late in life by Lecourt himself. In an 1990 article in France-Forum offering a retrospective on the origins of Article 49-3, including his own role, Lecourt asked whether its contribution to the effectiveness of French institutions may have come at an excessive price in terms of ‘democratic decline’.Footnote ¹⁰⁰ Lecourt answered that this question required deeper reflection, not only on Article 49-3 but on all texts whose legislative value derives from the implicit consent [assentiment implicite] of national representatives, and still more fundamentally on the legitimacy of all provisions which do not emanate from the nation itself [toute disposition qui n’émanerait pas de la nation elle-même]. Given the poor fit of traditional forms of parliamentary procedures to the volume and technical nature of problems facing modern legislatures, it was only under the benefit of a certain number of constitutional fictions that the law could be considered as an expression of the general will. So, Lecourt continued, under the pressure of events over the decades the practice had emerged covering all shades of consent, formal, implicit, spontaneous, resigned, silent or critical, in a ‘multifaceted form of acquiescence’.

Lecourt’s France-Forum article should be understood as linking his French and his European constitutional achievements. Although not explicitly referred to by name, European law must obviously be prominently included in any discussion – particularly with Lecourt as author – of ‘provisions which do not emanate from the nation itself’, and which obtain their national legislative value in France (and elsewhere) by implicit or silent consent (because of the direct effect doctrine). As we have seen, this movement of legislative outputs away from traditional procedures towards mechanisms that have only attenuated claims to national parliamentary consent, is the shared logic both of Article 49-3 and of the constitutional doctrines of European law. As Lindseth has convincingly argued, the legal development of the early European Economic Community cannot be separated from trends in postwar national constitutional development, in particular the explicit reinforcement of executive power at the expense of parliamentary factionalism.Footnote ¹⁰¹ In the person of Lecourt, indeed, the connection between constitutional developments at both European and national level is at its most obvious. Equipped already with the perspective that traditional parliamentary democracy as a means of rule-making was being increasingly overtaken by events, the ‘theoretician’ of Article 49-3 was well placed to become the great judge of the European Court of Justice.

To be sure, Lecourt was not in any sense the lone inventor or sole actor behind either of these constitutional developments. In the case of Article 49-3, the concept was nursed over the years by a group of MRP deputies, including Moisan, Coste-Floret and Lecourt, and its actual progress from the January 1958 proposal into Article 49-3 of the October 1958 Constitution depended on the support of, particularly, Gaillard, Pflimlin, Mollet, Debré, General de Gaulle, the majority of the National Assembly that voted to accept a version of Lecourt’s proposals, and, not least, the votes of the French people in the September 1958 referendum. As for the constitutional doctrines of European law, these had been nursed over the years by the Legal Service of the European Commission, under its director Gaudet, as well as by a variety of pro-European litigants and legal scholars, including Advocate General Lagrange, and their progress from ambitious legal claims to ECJ judgments depended on the support of, particularly, judges Lecourt and Trabucchi, the other ECJ judges and advocates general who supported the principles of Van Gend en Loos, influential judges’ assistants such as Gori, and, of course, after the Court’s own judgments, on the willingness of national judges, legal scholars, litigating lawyers, and state policy-makers to accept and make use of the new principles of European law over the 1960s and 1970s, and after.Footnote ¹⁰² In both cases, however, Lecourt played a key, perhaps irreplaceable, role in developing and promoting the ideas behind the new constitutional structures, and in spearheading their transformation into a practical reality.Footnote ¹⁰³

In short, the characterisation of the early ECJ by so much political science and legal scholarship as an anonymous and personality-less institution must at this stage be abandoned. To the major explanations of the development of the European legal order, which emphasise legal networking and judicial politics on the one hand, and facilitating economic and political conditions on the other, must now be added the fact that the Court of Justice at the time of its momentous legalisation of Europe’s international politics was driven by a dedicated ‘European’ who was also already one of postwar France’s boldest specialists in institutional reform and the entrenchment of executive power.

Finally, and on a wider methodological note, Lecourt’s background in constitutional design has been obscured only by the neglect of biographical study of ECJ judges.Footnote ¹⁰⁴ Even when obstacles appear daunting, however, the detailed study of judges’ prior political and constitutional commitments remains an unavoidable necessity, particularly for apex courts at times of radical novelty or disjuncture. It is judges that turn open-ended texts – national constitutions and treaties alike – into specific legal judgments. The example given by the study of the politics of the United States Supreme Court through judicial biography may therefore have much to contribute to the study of other powerful courts and tribunals, domestic and international, in Europe and around the world.