Towards the End of French-Style ‘Negative Parliamentarism’? Introductory Questions to the Study of the 2008–2009 Constitutional Reform

First published in French as ‘Vers la fin du « parlementarisme négatif » à la française ? Une problématique introductive à l’étude de la réforme constitutionnelle de 2008-2009’ (2011) 6, Jus Politcium.


The study of parliamentary assemblies has long suffered in France (and sometimes elsewhere) from many misunderstandings affecting the subtle issues that form its subject matter. Those issues depend primarily on the nature of law in general and constitutional law in particular. One must begin in this regard by recalling the inanity of an abstract dual vision claiming to distinguish between (constitutional) law ‘in books’ and (constitutional) law ‘in reality’.


The study of parliamentary assemblies has long suffered in France (and sometimes elsewhere) from many misunderstandings affecting the subtle issues that form its subject matter. Those issues depend primarily on the nature of law in general and constitutional law in particular. One must begin in this regard by recalling the inanity of an abstract dual vision claiming to distinguish between (constitutional) law ‘in books’ and (constitutional) law ‘in reality’. René Capitant and Charles Eisenmann, to cite two great scholars of recent times, some time ago (among others) underscored that it is obviously part of the jurist’s work to concern himself with the effectiveness and application of law (or better still, statements making normative claims).[1] In other words, it is important to realise that what is called ‘practice’ cannot be fully detached from law, that often it is even a moment in the making of the law itself[2] – which obviously does not mean that, in constitutional matters, all practice is law. This fairly elementary but fundamental point explains why it seems judicious to me, and others, to speak of ‘droit politique’ (ius politicum, constitutional law and politics or ‘polity law’). This means, too, that studies of parliamentary assemblies (as with any political institution) plainly need to be fuelled by exchanges between academics and practitioners.

Even so, French constitutional science is in urgent need of conceptualisations, especially in the matter of systems of government, an area in which doctrine has for too long been happy to make do with schemas that are too rough and ready, when not plain wrong. So a thorough analysis of Parliament and of the ‘new parliamentary law’ cannot dispense with conceptually situating the issues involved.

Parliamentary law, a subject matter that in appearance is essentially technical, as a part of constitutional law (Marcel Prélot) is underpinned by the same issues as constitutional law. Therefore it cannot be understood by ignoring the specific constitutional position of assemblies in a given state. More exactly, assemblies can never be meaningfully analysed independently of their structural relationship with the executive. In the case in hand, the principle of the political accountability of the government to parliament inevitably contains far-reaching implications that are not to be underestimated, even in assessing the technical parts of parliamentary law.

Examination of the consequences of the major constitutional revision of 23 July 2008 and its extensions (through the legislation accompanying and materialising it)[3] provides a good opportunity to try to test out the approach outlined here. After some two years in force, a provisional review of the reform can be made. Although this still means knowing how to evaluate the reform and determining what can reasonably be expected of it. Were the review to be undertaken in the absence of any prior consideration of the concepts and problems, it might prove overly naive and would unfailingly end up as a disappointment.

So it seems important to me to show that the 2008–2009 reforms fit into a constitutional panorama that is specific to the Fifth Republic, which I propose to characterise as ‘negative parliamentarism’ (or negative logic of parliamentarism).[4] It will then be possible to better understand the meaning of the formal changes made to the law in 2008–2009 and the effects they may have had on the actors of the Constitution.

We shall not dwell here on demonstrating that the system of government of the Fifth Republic can obviously be counted among the family of parliamentary regimes, even if some analysts (whether jurists, political scientists or others) believe this can be doubted because of the marked specificities of the French system. The eminent role played by the President of the Republic notwithstanding, the system is still structurally a parliamentary system. Not only are the formal legal rules in force in the system (especially the principle of the government’s political accountability to parliament) – and to be more specific it can be stated that its legal framework is that of renewed dualist parliamentarism – but its actual working (whatever its quirks and shortcomings) incontrovertibly also ties the Fifth Republic to the essential rationales of parliamentarism.

It is no less certain that French democracy has been characterised, since 1958, by an original system, which I have proposed to call a ‘presidentially captured’ parliamentary system (un système parlementaire à captation présidentielle).[5] The expression reflects the point that the President enjoys such a possibility of directing national policy because he taps into the force of parliamentarism (in this case, majority parliamentarism). Save in the special circumstances of cohabitation (power sharing), which is like an exception proving the rule, the parliamentary majority acknowledges the President as its leader. From that point on, it backs his every initiative (or almost) and the President can in this way ‘capture’ the force of majority parliamentarism for his own political benefit. The main official constitutional decision-making processes (legislative procedure, government internal discussions, regulatory power) are carried out, apparently, as in other Western parliamentary democracies, but they are actually instigated by the head of state and converge politically upon him. The French Prime Minister is seemingly in the same position as a British Prime Minister or German Chancellor, but in fact works for the account of presidential policy, not for his own political line alone. The President’s hold over things is legally elusive (apart, perhaps, from a large share of the power to make the major civil service and state sector appointments) since he has, in law, almost no power to directly govern affairs of state. In short, the President uses, or better still, he captures (or converts in the sense of tort law) these forces of parliamentary government. It is in this, constitutionally, that his true strength lies.


The parliamentary system of government is, as everyone knows, the most widespread form of government in the free world today. But behind its unity, this big family exhibits a degree of diversity.[6] Several variants of the parliamentary system can be made out, from many angles of analysis (formal structure, internal dynamics, parties, culture), making it all the more necessary to draw distinctions or even to try to draw up an obviously imperfect typology of them.

French doctrine currently proposes a number of distinctions to account for the variety of parliamentary systems. The essential pairings are: dualist/monist,[7] rationalised/non-rationalised, majority/non-majority.[8]

It will be observed that these distinctions can be combined with each other[9] and that they do not have the same aims. The rationalised character is relatively objective (it can be readily recognised by reading constitutional texts), although it may be of very varying intensity (not all relations between the executive and parliament are necessarily the subject of regulation; the minimum degree is codification of the principle that the government is accountable to parliament), but it is purely formal in that it says nothing about the working of the system. The dualist or monist character (which besides has intermediate degrees) may be either entirely formal (i.e. derived from the presence or absence of certain written legal rules) or substantive (when it seeks to make some account of the balance actually achieved in a given country between the head of state and the government, or even between the head of state and the parliament). Only the criterion of the majority or non majority character is derived exclusively from the actual working of a given system (regardless of the voting system to elect members of parliament), essentially by including the party system in the analysis.[10] But even enhanced in this way, it fails to fully account for the actual equilibrium and the overall sense, both legal and political (so to speak) of a given system.

The term ‘negative parliamentarism’ is sometimes used in Scandinavian constitutional doctrine, especially in Denmark and Norway,[11] to underscore that, in these two parliamentary monarchies constructed on classical dualism, the principle of presumed confidence prevails for the formation of the government. The head of the cabinet and ministers take office by royal appointment, without need for an explicit vote of approval from the parliament (which legally has only the option of no-confidence motions).[12] Being limited in scope, the arrangement goes no further and would probably not be quite adequate especially in Denmark for other aspects of relations between the executive and parliament because the formal intervention of the Danish Folketing is now required to validate many government initiatives (a point we shall return to).

The idea that parliament is reduced to conducting a ‘negative’ policy had been defended by Max Weber at the end of the First World War albeit in another context, that of the dualist monarchy with a non-parliamentary executive of the German Empire of 1871–1918.[13] Weber observed – and deplored – that the Reichstag under Wilhelm was, in the constitutional system developed by Bismarck, reduced to weakly controlling, from the outside a monarchical executive formed from outside its ranks; that its involvement in exercising the legislative function was limited to correcting or curbing the initiatives of the Chancellor of the Empire. The harmful consequences, in Weber’s view, of such a system led him, at the time, to plead for the setting up of a truly accountable government, that is a parliamentary government. Moreover, this ‘negative’ culture, which had deeply imbued the German elites, produced particularly perceptible effects in the institutionalisation and then the practice of parliamentarism under the Weimar Republic.[14]

The idea underpinned by these two references is worth taking up again, while widening and deepening it. It is possible to integrate in it, structurally, the examination of other rules and legal mechanisms (e.g. on the status of assemblies) and, politically, certain practices and even the constitutional culture. Only the combination of the three makes it possible to grasp the characters of a parliamentary system in particular and to apprehend its specific logic a little more finely.

It is striking to observe that different dynamics run through contemporary parliamentary systems. The principle of parliamentary government was born, as is well known, within the cultural context and the legal framework of liberal monarchy (whether limited, dualist or representative, to take up the subtle distinctions proposed by Philippe Lauvaux).[15] In its first viable historical phase, it has always been a matter for the assemblies (especially the elected chamber) of snatching a ‘right to concourse’, in short an initially modest but then growing involvement in the king’s government and the conduct of the country’s policy. While still essentially in the hands of the monarch (whose prerogative of appointing and dismissing ministers was still not seriously undermined), the executive, which remained the driving force of the state, largely retained control over its movements and enjoyed a vast power of autonomous regulation and as the principal instigator of formal legislation of limited scope.[16] The role of the assemblies was to control its action and that of the administration ex post (and even then in a limited manner), to approve the state budget, to discuss its initiatives (legislative or otherwise) while attempting to hold the government to account, in short to make government politically accountable and ultimately to secure from the King its removal or possible curbs on its policy. The assemblies’ role was therefore essentially reactive; they themselves did not determine the supreme political direction; they were, one might say, an opposition force.[17]

Their influence was able to grow with time. As supremacy of the elective principle and then finally of universal suffrage was gradually recognised, assemblies were at last to impose themselves politically on the monarchic executive. Progressively, it was accepted that a country’s policy and therefore the government tasked with leading it, should be determined essentially by the political orientation of the parliament (more exactly the parliamentary majority). The end point of this principle was the de facto or de iure capacity of the assemblies to impose on the head of state the persons in charge of government, which Bagehot identified for England[18] from 1861 and which was shortly (in 1865) to be named the ‘electoral function’ of parliament. Where the evolution was not held in check (further to the inability of members of parliament to articulate any clear political will or further to an authoritarian turn of the regime, as in Weimar Germany), it flourished either through a set of constitutional conventions enshrining the new balance of forces beyond the letter of unchanged strict law (as in almost all European liberal monarchies) or by formal legal crystallisation of new constitutional rules (notably in republics after 1918).

This fundamental change cannot be reduced to the transition from dualist to monist parliamentarism (handy and generally useful although somewhat reductive formulas). It is possible to go further and consider that the mainstream historical trend of the parliamentary system of government, in almost all countries where it was practised over a long period, led to a genuine paradigm shift: representative assemblies went from being institutions for limiting what remained fairly largely autonomous governmental power to being the main source of legitimacy of that power, with this new position reflecting, at bottom, the primacy of universal suffrage, that is of the sovereign people. Not only did the formation and maintaining of the organ of government depend henceforth almost solely on the assemblies (the hangover of a residual role for the head of state being essentially construed as the sign of sporadic or chronic malfunctions of a particular system, notably in the event of instability of the party system and of parliamentary forces) but in addition the requirement that the essential part of government action should be legitimated by the assemblies was considerably extended and progressively formalised. It came to be considered that most of the measures the government counted on taking were to take the form of a parliamentary statute (the initially limited domain of which came to encompass all matters) or indeed of another form of parliamentary approval (vote for budget, vote authorising treaty ratification, vote approving a declaration of war and then even the authorisation to commit armed forces, etc.). The purpose was to reflect the idea that government was to rest positively on the trust and support of the assemblies. This tendency is obvious in young democracies, especially parliamentary republics arising from the overthrow of a more or less authoritarian regime (thus, for example, old or new states in central and eastern Europe after 1918); it is scarcely less clear in most of the older regimes, including in monarchic form, generally marked by less discontinuity of constitutional law (as in the northern or Benelux countries). This has been reflected, apart from by a reinforcement of status,[19] by a marked increase in the legal competence of parliamentary assemblies guaranteed by constitutional texts.

In the first place, new processes for forming governments were dreamt up. Although the legal competence for the head of state (monarch or president of the republic, which was usually simply the republican transposition of the monarch, the typical example being the president of the Third French Republic) to appoint ministers was maintained, this called for a prior parliamentary vote (for example in the 1937 Irish Constitution), investiture vote (1946 French Constitution) or an initial parliamentary vote of confidence (e.g. 1947 Italian Constitution or the 1975 Greek Constitution). Sometimes things went further by instituting the direct election of the prime minister[20] by the parliament (1920 Austrian Federal Constitution, 1946 Constitution of Japan, 1949 Basic Law of the FRG, 1989 version of the Constitution of Hungary). Whether requiring an absolute or a relative majority, this new constitutional mechanism attested explicitly to the recognition of the parliament’s electoral function,[21] instead of which it has remained informal in some monarchies (United Kingdom and most of its former dominions): the wish was thus for the government to be legitimated from the outset and positively by the assembly (and sometimes also by the second house as in Italy). This was to attempt to consolidate, to give a positive basis to the principle of parliamentary government.[22] In a sense, that could have been enough for it to flourish, while leaving considerable leeway for the cabinet set up in tune with a majority.

Yet it is observed that most constitutions have, over the course of time, extended the legal powers of assemblies, either by explicit wording in the actual constitutional text or (when they were spreading, that is, from the second half of the twentieth century) through the case law of constitutional courts: on legislative and budgetary matters first of all, for which restrictions on assembly involvement disappeared.[23] Then in the diplomatic and military domains, the areas of Lockean federative power, where the assent of the houses was from then on required to authorise ratification of the main treaties, to declare war or to conclude peace[24] or, more recently, to authorise the engagement (or its extension) of armed forces abroad.[25] The same logic is at work when a constitutional court that does not shy from law-making, like that of federal Germany, imposes a ‘parliamentary reserve’ (Parlamentsvorbehalt) to require parliamentary legitimation of government decisions on European matters.[26] Likewise too for the designation of the holders of certain eminent public offices and positions, for which the pure and simple election or at least approval by parliament of the proposed appointment by the executive (or other organ) is required.[27] The idea underlying the extension of the powers of assemblies is to force the government, although it is itself already, in principle, a parliamentary cabinet, to have its initiatives and the political measures it deems it must take made legitimate. Indeed, these initially formal powers are both the opportunity for the majority to show its agreement with the ministry (if need be after negotiation and compromise) and, for the opposition to publicly voice any criticism it may have. As such, it is a means of ensuring government through public discussion.

The fact remains that alongside the provisions of constitutional law, which form a general structural framework, consideration must be given to institutional practices and the behaviour of actors if one is to claim to comprehend the actual balance of a system of government. From this point of view, such practices and the technical provisions may not always be fully matched. It can be observed that, in most classical dualist monarchies, constitutional conventions and other political practices often have a substantially ‘positive’ impact on techniques which have generally remained those of the logic of negative parliamentarism. This is true of the United Kingdom, because of the very close merging of powers between the cabinet and the House of Commons and the (albeit imperfect) two-party majority system: the cabinet’s technical freedom of movement is offset by more or less informal mechanisms allowing symbiosis between the cabinet and its parliamentary majority. This seems also to have been the case, mutatis mutandis, in the Netherlands for some thirty years, after a long period of relative distance between the cabinet and the Estates General (promoted by an unstable multiparty system and minority cabinets). It is rarer, conversely, for a system technically structured according to the logic of positive parliamentarism to move radically away from it in political behaviour, although it would be quite conceivable. In truth, only meticulous analyses, over the long term, could establish exactly what is what in each country. However, an essentially positive culture can be identified in practice by examining parliament’s actual exercise of the electoral function: it is the major criterion for telling whether the government is fully tied and bound politically to a majority within the parliament. From this point of view, the practice of governments of technocrats, which has not entirely vanished from modern party democracy, illustrates a trend curbing the burgeoning logic of positive parliamentarism. But there are still many other types of practices and usual behaviours, not covered by law, expressing one or other of the logics of parliamentarism.[28]

Although negative logic was characteristic of the early days of parliamentarism and has tended to fade away in a growing number of countries to be replaced by positive logic, it should be added that this has often happened gradually and has left intermediate stages in place in that a number of countries mix the two logics de facto or de iure. Obviously it would be unreasonable to think that the ‘positive’ variant could be achieved in the pure state. As a rough and ready guide one can imagine a classification of parliamentary systems by this criterion as a scale at the two ends of which we have an essentially negative logic and an essentially positive logic. Concretely, countries having practised the system of parliamentary government at its beginnings (Great Britain in the eighteenth and nineteenth centuries, other European parliamentary monarchies in the nineteenth centuries) should be ranked at the negative pole, and contemporary states at the positive end, with a whole array of shadings between the two.

The Federal Republic of Germany since 1949 embodies the ‘positive’ model fairly well; most of the republics practising the monist variant (Hungary, Greece, Italy, etc.) are close to it. Conversely, regimes having conserved a formally dualist constitutional structure are generally more remote from it: Finland up until the reforms of the 1990s, Poland, etc. Denmark is a less clear-cut case: the 1953 Constitution maintained a classical dualist framework, but the parliament’s reserve powers are substantial and have been markedly increasing for several decades, with the result that, apart from the procedure for forming the cabinet, the country is more one of positive logic.

Thus the concepts of negative and positive parliamentarism represent two extremes of the realisation of the principle of parliamentary government, depending on how great a requirement is made of the anchoring of the government with respect to parliament.[29] Negative parliamentarism (or the negative logic of parliamentarism) requires only weak legitimation of the government and its actions by parliament, with the result that parliament has only a negative influence; positive parliamentarism, on the contrary, imposes strong legal and political dependence of the executive on the parliament, whose influence on the conduct of national policy proves relatively strong. This distinction depends on the combination of purely legal, structural factors and of more or less contingent political factors that either help or hinder the potentialities.


It is common knowledge that the advent of the Fifth Republic was a major legal and political, ideological and practical break point in French constitutional tradition. Nor did the promoters of the 1958 Constitution hide the fact. It was their clear intent to break with the system of ‘parliamentary sovereignty’ that had prevailed under the Third and Fourth Republics, a system in which the assemblies were de iure and de facto masters of the part they wanted to play within the constitutional order. Without entirely abandoning the parliamentary basis of the system of government,[30] it was therefore thought necessary to ‘sanitise it’ (l’assainir), as Raymond Janot put it.[31] This undertaking was reflected by two sets of changes to the constitutional framework: one, essentially technical and precise – that was soon connected with the formula invented elsewhere of rationalisation of parliamentarism – affected the standing and powers of the assemblies (and, by repercussion, directly the government itself as an organ); the other also (but in part only) legal but less precise as to its scope, sought to establish something like a ‘state power’ (to take up Georges Burdeau’s expression) around the President of the Republic, a power that was largely out of reach of the assemblies. These two relatively independent sides (they could have been adopted one without the other) came together to produce a radical change in the spirit and the outcome of the French constitutional system. To explain this system, a distinction should be made between the realm of constitutional law and the realm of the system of government.

2.1. The Realm of Constitutional Law: One-Way Rationalisation of Parliamentarism

The rationalisation of parliamentarism initially (as conceptualised by Boris Mirkine-Guetzévitch on the basis of new constitutions of central and eastern Europe in the inter-war years[32]) means a codification, a legal shaping of relations between the executive and the assemblies which until then had relied on unwritten rules and practices. In itself, it is an essentially neutral formula in that it can be aimed equally at consolidating the constitutional position of assemblies or tempering their means of action on the government.

In the French case, it was used methodically in 1958 to ring-fence the assemblies in a limited way and ensure as much freedom of movement as possible for the executive. Its potentialities have, besides, been systematically exploited (by the government, aided and abetted by the Constitutional Council). In this sense, it has worked in one direction to the detriment of parliament.

From all of the statements of written law (in the formal Constitution and its implementing regulations) newly established in 1958–1959 and their stark application, it can be observed that the ‘negative’ logic of parliamentarism was largely instigated in French constitutional law. The prevailing principle from then on was that the parliament was prohibited from doing whatever it was not expressly authorised to do.

This was the case first of all for its organisation:

  • the assemblies no longer sit when they wish but only in two three-monthly ordinary sessions per year.[33]
  • the internal regulation of each assembly (as besides that of the two joined in Congress) is overseen by the Constitutional Council (in contrast to the traditional principle of the internal autonomy of parliaments).
  • the parliamentary agenda and work is decided as a matter of priority by the government (in practice, this priority has rapidly verged on a near monopoly).
  • the assemblies may organise themselves as they wish: the Constitution limits the number of standing committees to six and the status like the investigating powers of the committees are very limited (by a statute initially drawn up by the government).[34]

Likewise, in their acts and decisions, the representative assemblies may express themselves using strictly limited procedures only:

  • on legislative matters, the assemblies do admittedly retain the power to propose legislation (but these proposals are subject to strict limitations as to their material and financial admissibility), to discuss it[35] and pass it, but through various prerogatives the government directs the essential part of the legislative procedure (without counting that it holds greater regulatory power than before).[36]
  • the government may make a vote on any statute into a vote of confidence in the National Assembly (art. 49(3)) so as to force members not to oppose its bills or to do so at the cost of bringing the cabinet down. This ingenious article in itself encapsulates the negative logic of parliamentarism since there is a presumption in the favour of the government, and it is for the members of parliament to overturn that presumption…by overturning the government.
  • the assemblies can no longer pass resolutions on general policy.[37]
  • the electoral function of parliament is no longer positively recognised: the government, appointed by the president (art. 8), – at the end of an interpretation imposed by the executive and finally accepted by the members of parliament[38] – does not require any investiture or vote of confidence from the National Assembly; the National Assembly may only take the initiative of passing a motion of censure of the government (art. 49). The principle of ‘presumed confidence’ is enshrined in this way.
  • members of the assembly or senators joining the government cease to exercise their parliamentary mandate and do not recover it automatically upon leaving office (arts 23 and 25C). This measure, breaking with French tradition and the spirit of parliamentary government, was designed to distance ministers culturally from the assemblies.

Finally, the assemblies can only:

  • discuss and vote on legislation (art. 45),
  • ask written and oral question of ministers, with no legal consequences (art. 48),
  • pass a motion of censure against the government (art. 49(2)), a decisive weapon but one that cannot be wielded every day.

Thus the rules of law of the Constitution (including parliamentary law), that is, of positive law, converged very clearly in the direction of negative logic. Yet, they were not in themselves to entail unfailingly, in a strict cause-and-effect sequence, a real abasement of the assemblies. While containing powerful constraining potential, most of the legal rules for rationalising parliamentarism remained ‘permissive’ rules, leaving constitutional actors leeway to use those rules, which they could have done moderately. This was not to be, as is well known, (this was the case of the government and the Constitutional Council, the prime organs for materialising constitutional statements and which acted as secondary constitutional bodies, accentuating through their decisions and attitudes the tendency to restrict the assemblies). Moreover, the negative logic of parliamentarism was indeed deployed from the early years of the regime and finally lastingly installed because all of the interpretations, institutional practices, political decisions and discourses on the Constitution converged in favour of an effective system of government that instrumentalised the acting forces of parliamentarism.

2.2. In the Realm of the System of Government Practised: Presidentially Captured Parliamentarism

Most observers (constitutionalist doctrine included) of constitutional life readily believe that a country’s system of government results logically from the ‘Constitution’, understood as constitutional legislation. In this, they fall victim to a realist ontology (as Michel Troper might say), consisting in forging straightforward linkages of cause and effect between what they identify as positive law (in fact, the plain statements of written law of the Constitution – but these only become positive and effective legal norms after being materialised, interpreted and implemented) and the general equilibrium achieved in practice. It is thus very often claimed that the ‘Constitution’, apart from the impressive arrangement evoked earlier of rationalisation of Parliament and of its relations with the executive, establishes a presidentialist system (to use a handy although terribly vague term here). And that this presidentialist system, being by nature different from what a parliamentary system should ‘normally’ be, it is unsurprising that the parliament of the Fifth Republic is particularly weak.

In actual fact, not only can such simple causal links not be seriously asserted, as just said, but in addition, the ‘presidentialist’ character of the Fifth Republic is relatively independent of rationalisation. It so happens (it was not inevitable) that in practice, presidential power of impetus and direction actually has been deployed since the years 1959–1962 and took root in the ensuing decades, even if it was neutralised from time to time by cohabitations. This course taken by the Fifth Republic is too familiar to need any detailed presentation here. Even so, it is important to realise that it is this dimension of the system of government, what I call ‘presidential capture’, that is decisive for the blossoming of the negative logic of French parliamentarism.

The presidentialist ideology – and finally the growing consent within political elites at least in the main parties of government – explains, admittedly, the methodical materialisation from 1959 onwards of constitutional statements in the sense of effective restriction of the assemblies by virtue of the law of the Constitution. But, more still, it is what one might call the presidentialist habitus that has radically accentuated and even geared down the effect of legal constraints weighing on the assemblies and contributing to the deployment of the negative logic of parliamentarism.

It all begins with a point of pure fact and not in the least of a legal rule: the sheer determination of the first President of the Fifth Republic, de Gaulle, which was taken up by all his successors for their own account, to exert supreme control over national policy. The first step in achieving this ambition was (and remains) obviously to form a government he has a grip on and first, to personally choose the Prime Minister.[39] Even so, political means and material resources were still needed. Technically, this operation is facilitated by the legal arrangements for appointment set out in articles 8 and 49 of the Constitution: as is known, the rule was imposed that the appointment of the Prime Minister by non countersigned presidential decree suffices to invest him fully in his powers, that article 49(1) has finally not been construed as requiring a prior vote of confidence. Yet this principle of presumed confidence did not imply in itself the substantially (and not merely formally) dualistic character of the regime; this character derives from the practice of the presidents and accepted by successive prime ministers and finally by all members of parliament (more specifically of the successive parliamentary majorities).[40] But it was the condition for it in practice.

The technical principle of presumed confidence is found in many Western parliamentary constitutions (without speaking of monarchies, where the role of the ruler is now – saving exceptions – conventionally neutralised from the outset, one can cite a republic such as Austria since 1929) without for all that implying a personal choice of the head of state for appointing the leader of the cabinet. The French case differs because the successive presidents have sought to exercise this power and, since de Gaulle, this claim has been progressively accepted (by the supporters of the head of state,[41] then by his opponents) and is therefore considered a true constitutional ‘norm’ to the point of holding pride of place among the string of reverences towards the presidential office.

Of course, such a substantive interpretation of a legal power (art. 8(1) and a contrario art. 49(1)) as a component of a system having its own logic (presidentialism) could probably not have imposed itself without the advent, in the National Assembly, of majorities that were well-disciplined and above all devoted to the president, who, from 1962, blocked the ‘countermeasures’ parliament might have raised against the president’s interpretation.[42]

Similarly, the recognition of a real presidential prerogative to choose the prime minister entailed, despite the wording of article 8(1) second sentence of the Constitution, that of revoking or, more exactly, as Pierre Avril pointed out, of the accountability of the prime minister to the head of state.[43] What one might characterise as a true convention of the constitution is obviously necessary to maintaining the presidentialist logic and has only been able to establish itself by what can only be called the feeling of political subordination of prime ministers with respect to the president, it being further specified that there was never any doubt about the fact that the parliamentary majority would tacitly acquiesce to this convention whenever occasion arose.

Once the principle was established of a cabinet actually ‘proceeding’ from presidential volition, the composition of governments came to fit this logic, even if some account had to be taken of the parliamentary make-up (notice in passing that de Gaulle himself gradually rallied – or became resigned to? – a choice of ministers who were ever less ‘technicians’ or without party allegiance).

From that point on, the phenomenon of ‘presidential capture’ could be put in place, tying back in, in a way, with the balance of dualist parliamentarism of yesteryear: the ‘King’s government’ acts, makes decisions and legal instruments for the account of the monarch, with more or less autonomy depending on the latitude allowed by the head of state himself. Formed by royal will, it governs in accordance with the more or less direct instructions of that volition, relying on a parliamentary majority that essentially fulfils a support function and does not think of itself as much as a driving force as a force of reaction to the executive’s initiatives.[44] Strict party discipline, combined with reverence for presidents has promoted this passivity.

Probably, if one puts aside for a moment the French specificity of the role of the President, it should be emphasised that the French assemblies encounter the same constraints as their counterparts in all modern parliamentary regimes (strict duality between majority and opposition, strict voting discipline of members, overseeing of the member’s work by the political group and the predominant role played by the executive in deliberations and the decision making process in the parliament). But there is also, in the French case, a propensity of members of parliament to feel satisfied with their condition. They have interiorised their abasement. Apart from the lack of dynamism of the opposition (which has often given the impression of preferring obstruction to ground work), it is above all the self-restraint of the members of the majority with regard to the executive that characterises French parliamentarism. These members are often too docile and do not demand enough from the government for their support.

It would be illusory to think:

  • that a parliamentary assembly might escape from relatively stable internal political structuring into a majority and minority (the whole history of deliberative assemblies shows this plainly and the case of the United States, a regime where the executive is not politically accountable to Congress, confirms it);
  • that a parliamentary majority, in principle made up by reference to a certain political line and therefore in a system where the principle of accountable government prevails, by reference to a cabinet in particular, can make its own determinations regardless of the government’s positions. Every parliamentary majority in parliamentary systems of government is to some extent in the political grip of the cabinet. The government needs the discipline of its majority, without which it cannot govern lastingly. It is rare for a parliamentary majority to escape the complaint of docility, and primarily from its opponents. The question is to what extent and with what prospects.

The comparative political weakness of parliamentary majorities under the Fifth Republic remains a striking feature by comparison with other countries practising parliamentary government. The reasons for this state of affairs are not exclusively legal ones (all the more so because sporadic reforms were already adopted before 2008). It is more profoundly a question of a change of parliamentary culture that the Fifth Republic caused. It does indeed seem that this cultural change has developed and become rooted for a structural reason to do with the fact that the National Assembly has lost the monopoly of legitimising government that it held before 1958. Since it has gradually been accepted that the President of the Republic, himself elected by direct universal suffrage (but contrary to members of parliament, by the whole of the electoral body), was the first purveyor of government legitimacy, the role of parliament in the government’s existence has moved to the background and is confined to supplementary, almost accessory legitimation.[45] While remaining fundamentally essential, the support of a majority of the Assembly is, under the ‘presidential capture’ system of government, no longer the only crucial point for the existence and continuation of the government. The President in some sense over-determines the relation between the government and the majority. The dialectic inherent in this relationship is complicated, in the French case, by the presidential factor, which explains that the majority should be ‘slave’ in the cybernetic sense of the world: it is controlled from outside.[46]

It can thus be understood why and how the negative logic of parliamentarism has been able to develop in France: the ‘orthopaedic’ rationalisation (Marcel Prélot) was amplified from the outset by the return to the dualist variant – renewed (because supported by universal suffrage) – of parliamentarism which, against a background of majority domination and presidentialisation, has been stabilised (except during cohabitation) by presidential capture.


It is common knowledge that the straitjacket imposed on the assemblies by the 1958 Constitution was slackened somewhat especially in 1995 by the introduction of the ‘single session’ of nine months (instead of two three-month sittings) and by the creation of the ‘parliamentary window (or ‘niche’)’ allowing each assembly to enjoy a priority over the agenda for one session per month (art. 48 C(3)).[47] Previously, the stringent limitation of the number of standing committees had been doubly loosened: by the creation, as from 1972,[48] by legislative means of ‘delegations’ or ‘parliamentary offices’ either specific to each assembly or common as required, the number of which has gradually increased;[49] but also by the practice of temporary parliamentary missions inaugurated in 1972 in the Senate and multiplied since then. Similarly, it could be pointed out that the government’s precedence over the agenda was, in practice, substantially restricted by various constraints.[50]

Having set itself the general objective of ‘rebalancing’ and ‘modernising’ the institutions of the Fifth Republic, the think tank presided by former Prime Minister Edouard Balladur in 2007, which directly inspired the essential part of the constitutional revision of 23 July 2008, particularly highlighted the necessity to ‘strengthen parliament’. This sweeping – and somewhat hackneyed – formula, underpinned by a rather brief assessment of the failings of French parliamentarism (it will be recalled that the committee had barely three and a half months to complete its report)[51] might make one fearful of a wrong or clumsy diagnosis of the position of the parliament of the Fifth Republic – all the more so because the committee did not claim to radically challenge the ‘presidentialist’ character of the system of government. In truth, it must be acknowledged that the Committee showed a degree of daring in its proposals, especially with respect to the Parliament, the whole going well beyond the avenues suggested by the President of the Republic. Most of its proposals were taken up by the legi-constituent assembly (or by the materialising organs), which sometimes went even further.

We shall evoke the main innovations of constitutional law[52] and ask whether they are of a nature to curb the negative logic of French parliamentarism.

3.1. The Status and Organisation of Parliamentary Assemblies

The number of standing committees each assembly is authorised to form was increased to eight (from six) (art 43 (1 C)). In point of fact, only the National Assembly took advantage of this extension straight away, while the Senate first refrained from creating new committees. Further to the three-yearly renewal of September 2011 – and the political change – the formation of one or two additional committees (one of which would be tasked with ‘sustainable development’) is contemplated. In any event, this extension is measured to say the least and makes little change to the fact that the assemblies are not entirely free to organise themselves as they wish.

The agenda for parliamentary sessions has, on the other hand, been much reorganised: the principle is now that the parliament sets its agenda freely (within the bounds of the number of days’ sessions laid down by article 28 C) (Art. 48 C (1) sentence 1), saving exceptions – which are substantial, it is true. Finally, only a big half (and not virtually all) will be determined primarily by the government, the remainder being left for the assemblies. Moreover, one session per month is reserved for opposition groups and minority groups. This point of the reform, which markedly amplifies the change made in 1995 is not insubstantial, even if one is sceptical about how useful it really is to favour the legislative initiative of members of parliament. It will be noted by the way that the legi-constituent assembly did not think it was waiving being directive since it prescribes that ‘one week’s sitting out of four is reserved as a priority for scrutiny of government and evaluation of public policy’ (art. 48 C (4)).

3.2. Scrutiny of the Executive

Parliament can henceforth, through its standing committees, object to the main appointments to the major state positions (foremost for the members of the Constitutional Council, but also for the chairs of the independent administrative authorities, etc.)[53] (art. 13 C (5)). This remarkable innovation is, however, neither an investiture nor a positive confirmation but a power of veto with a qualified majority (to block the appointment the tally of ‘no’ votes in one or both committees, as the case may be, must amount to three-fifths of the votes cast). Legally, this system designed for the executive to maintain the power of appointment is indeed a part of negative logic: members of parliament must prove they are numerous enough to explicitly object to an appointment. However, the mechanism might have dissuasive power and so prompt the President of the Republic to discretely check beforehand that there are not too many opponents on his own side (it seems this preventive effect has already borne fruit in practice).[54] Two serious objections can still be levelled at a system of the kind: it is disputable whether one can put on an equal footing the position of chair or director of public companies or bodies (the national centre for scientific research or the meteorological office) and positions in important constitutional institutions, especially the Defender of Rights or members of the Constitutional Council, as if they were appointments of the same kind.[55] Moreover, what is to be made of the sore situation in which parliamentary committees have a straight majority but not a three-fifths majority vote against? The President could then still appoint someone whose designation the Parliament has solemnly disapproved of. It seems difficult not to see in this a culture that is disdainful of the legitimacy procured by the representative assemblies.

The assemblies recover the right to vote on resolutions about general policy (art. 34-1 C). Nearly twenty years after their reintroduction limited to drafting European Community legislative instruments (art. 88-4 C), this restoration comes with strict conditions reminiscent of the stringent case law of the Constitutional Council in 1959. In specifying that ‘any draft resolution, whose adoption or rejection would be considered by the Government as an issue of confidence, or which contained an injunction to the Government, shall be inadmissible and may not be included on the agenda’, article 34-1 C (2) ( clearly keeps a lock on the spirit of initial negative parliamentarism: the French parliament can only express itself through resolutions to the extent that the executive does not deny it that right. Some might argue this is a wise precaution because unbridled use of resolutions might weaken the government’s political position and even drain of its substance the closely rationalised mechanism for challenging the government’s political accountability put in place by articles 49 and 50 C (the government possibly thinking itself compelled to put its political accountability to the test outside of official procedures to thwart a draft resolution that does not suit it). It must nonetheless be pointed out that the ‘material’ limit of article 34-1 C (2) contains a simplistic vision of the political relationship of trust between government and parliament. This relationship cannot be reduced to direct public votes on the government’s existence but is conveyed by the permanent intention (even if sometimes steeped in acrimony, tension and conflict) to work together in leading the country. In other words, confidence is an intrinsically political relationship which is nourished primarily by what is not always public. If a draft resolution containing no injunction upon the government and not explicitly directed at disapproving of its conduct sets out a position (on some topic or other) that is uncomfortable for the executive, will it be accepted that the government can hold it inadmissible? No provision is made in the event of disagreement over the interpretation of this article 34-1 C (2).[56] If the government does not raise the objection of inadmissibility, will it not be tempted to stake its continuation unofficially on members of parliament (principally those of its majority) to defeat the resolution? It can be seen from this that parliamentarism can only be rationalised to a certain extent, since the relation of confidence between the chambers and the cabinet cannot be fully ‘apprehended’ by law. The limitation posited in article 34 C (2) is therefore somewhat artificial and superfluous. It attests at any rate to persistent distrust of parliament.

In any event, these considerations can still be refined by observing that, in practice, and notably in a political configuration of ‘majority’ parliamentarism, resolutions are usually a means favoured by the government itself for allowing its majority to express what it cannot (or will not) itself say explicitly, and that the government is sometimes relieved to be able to rely on with respect to public opinion, interest groups or foreign governments. There is, as it were, an agreed sharing of roles of the majority bloc with respect to the outside audience.

The executive now has a duty to inform parliament of the dispatch of armed forces abroad. If the operation lasts for more than four months, approval by a public vote of parliament is required (art. 35 C (3)). This introduces, on certain terms, a genuine ‘reserve of parliament’, a new formal power of approval of a decision by the executive on diplomatic and military matters. Experience had cruelly pointed up this void in the French Constitution, which is ill-adapted to situations that have multiplied in the last twenty years or so of military operations conducted as part of concerted actions by members of international defence organisations, especially the United Nations. Comparative law has come to provide specimen solutions. But it can be noted here again that the legi-constituent assembly preferred to advance modestly by limiting parliamentary power to operations lasting more than four months (the Balladur Committee suggested three months).[57] By way of comparison, the agreement of Germany’s Bundestag is required from the outset for any operation by the federal army outside of the country and promptly in the event of absolute emergency.[58] In the French case, positive logic is therefore only partially recognised since parliamentary approbation is not a prior condition and is required only after a relatively long period. Despite these restrictions, it is incontrovertibly a step forward for parliamentary scrutiny.

Rules on commissions of enquiry were not – rather surprisingly – the subject of sustained attention from the Balladur Committee (whereas it claimed to want to enhance parliamentary control), no more than from the legi-constituent assembly, which merely constitutionalised such commissions (art. 51-2 C) without that affecting the terms on which they are formed and operate. Yet an important further innovation of the National Assembly’s rules deserves to be pointed out: the opposition will have greater facility to obtain the creation of a commission of enquiry. Henceforth, each chair of an opposition group or minority group may ask, once per ordinary session, for a debate on a proposed resolution to create a commission of enquiry to be listed on the agenda of a sitting of the one week per month reserved for scrutiny (art. 141 (2) RAN).[59] But above all, at the outcome of debate, ‘only members not in favour of creating the commission of enquiry take part in the vote. The application for creation (…) may be rejected by a three-fifths majority of the members of the Assembly’ (art. 141 (3) RAN). Although this is still a long way short of a constitutionally guaranteed unconditional right of the minority as in some neighbouring countries (Germany, Portugal and Greece, in particular), the new arrangement in principle leaves good chances for the opposition (which usually has a stake in the investigation being brought) to see its application succeed. It can be observed that it is for a strengthened majority against the principle of an enquiry to overturn the presumption. However, the situation remains unsatisfactory: one must also count with the power of the Government[60] or even of the president of the assembly (after opinion of the bureau) to refuse to set down the proposal for reasons of purported inadmissibility of the purpose of the enquiry (art. 138 (2) RAN)[61] or again the option of the majority to modify the purpose of the enquiry to such an extent that its initiators consider it has been voided of its substance.[62] In both instances, the sponsors of the application for the enquiry have no resort, so that the right hangs on the willingness of the majority of the Assembly. It is hard not to see this as a missed opportunity to provide the opposition with a real instrument of scrutiny.

Lastly, it will be noted that although generally the constitutional revision of 2008–2009 includes some advances in the recognition of the opposition and the instruments it can wield in the parliamentary process, notable restrictions remain, as seen with respect to commissions of enquiry. Likewise, the option provided for in article 50-1 C of organising a debate in public sitting on a government declaration, either at the government’s initiative or at the request of a parliamentary opposition group or minority group falls well short of what would have been desirable in order to enable the opposition to force the government to explain itself on any particular subject. The terms of article 50-1 C (obviously confirmed by the rules of the assemblies,[63] which could not add a constraint where the formal Constitution did not impose any) are such that the government remains free to decline the debate called for by the opposition, just as it may conclude any debate by an indicative vote.[64] This means that if the government shrank from this, the opposition would be bound to use the one reserved sitting per month to force the debate, which is plainly absurd. Here again, the reform seems to be too faint-hearted.

3.3. Legislative Procedure

An enormous mirage, as old as modern representative government itself, continues to loom on the horizon of common opinion, whether that of citizens, (supposedly) ‘enlightened’ observers or even many of the governing class themselves, in France as elsewhere: the idea that the main purpose of parliamentary assemblies is to exercise the legislative function and to do so as autonomously as possible. Despite all of the features of law or fact demonstrating how inane it is, especially in parliamentary systems of government, this idea persists. Accordingly, it is understandable that the introduction of new constitutional rules framing the legislative activity of the parliament in 1958 might have been considered a step backward. However, it was fundamentally merely attuning (sometimes over stringently, it is true) the role of assemblies in the logic of a modern parliamentary system, which implies the leading role of accountable government in the design and preparation of the instrument of government that modern law necessarily is. The rules introduced in 1958 for legislation were as much a part as the others of the negative logic of parliamentarism.

In any case, and without emphasising further here the illusions of the reformers of 2008–2009 as to the restoration of an essentially autonomous exercise of the legislative function by the parliament, it is possible to examine the main innovations in the perspective of negative parliamentarism.

The government will no longer be able to stake its accountability officially on the vote of any bill: henceforth, it will only be able to do so for a finance bill or a bill for financing social security and for a single text (government or private member’s bill) per session (art. 49 C (3). This is in law one of the most striking changes in the 2008 constitutional revision, which gives satisfaction to the many critical opinions on this apposite – but particularly ingenious – instrument of negative parliamentarism. It remains but has been considerably blunted. Arguments in various senses could support criticism or on the contrary approbation of this reform. In any event, the 49-3 (motion of censure) can no longer be an ordinary means of government and that is a distancing feature with regard to negative logic.

Among the many innovations on legislative procedure and which have not been detailed here, three provisions should be indicated that distance French parliamentarism from negative logic.

First of all, discussion in public sittings will engage (exceptions apart)[65] on the text arising from the works of the commission and no longer on the government text (art. 42 C (1)). It is a considerable reform since it will enhance committee work, which has long become essential in modern legislation. It will limit ministers’ offhandedness with respect to the work of the committee members and force them to be more careful in their dialogue with members of parliament. Once again, the ‘presumption’ in favour of the government is reversed: it will be up to the government to positively convince the members of the assembly or senators to follow it.

Then, there is the possibility for the Conferences of Presidents of the two assemblies to jointly oppose the application of the fast-track procedure (formerly the emergency procedure) wanted by the government for the discussion of texts) (art. 45 C (2)). This is, true, a veto of a government decision – in other words a reaction – but the significant point lies in the fact that the fast-track procedure can no longer be imposed on the assemblies unilaterally.

Lastly, the presidents of the two assemblies may convene a joint commission to settle disagreements between the two chambers over a bill (art. 45 (2) whereas previously the prime minister alone could decide on this. The head of the government does however maintain the monopoly to bring before the assemblies the conclusion of such a commission and, as the case may be, to ask the national assembly to make a final ruling.

3.4. The Return to Parliament of Former Ministers

Without going so far as to return to the incompatibility between government functions and parliamentary mandate (art. 23 C (1)), the 2008 revision made the regime more flexible: former ministers taken from parliament may automatically recover their parliamentary seat a month after leaving their governmental post (art. 25 C and organic law no. 2009-38 of 13 January 2009). One of the emblematic measures of the change in 1958 has therefore been toned down. It is common knowledge that it had rapidly failed to comply with what General de Gaulle’s had initially intended, since political figures in the government did not relinquish their quest for electoral legitimacy. By allowing the natural return to parliament of former ministers, the new rule should contribute to bringing the executive and parliament culturally closer as it will facilitate smooth transitions between the two centres of the government majority. As such, it is also a (small) step away from negative parliamentarism. What would have been truly audacious would have been to completely eliminate the incompatibility and even to require that at least a majority of ministers should be members of parliament as do some constitutions inspired by the Westminster model.[66] It was obviously illusory to expect such audacity from political personnel so imbued with the spirit of negative parliamentarism.


Has the 2008 constitutional revision removed the French constitutional system from the ‘negative’ logic of parliamentarism? The answer cannot be a clear-cut one.

First because evaluating the effects of a reform of constitutional law on this scale requires hindsight. It is obvious that those involved (members of parliament, ministers, ministerial advisors, parliamentary assistants and constitutional judges, especially) will need time to measure the scope of the changes, to use them, discover their potential and, as the case may be, ward off attempts to neutralise their potential (especially by the government in office, whatever its colour, and its political supporters insofar as the new powers of parliament are liable to hamper whoever is in power). Although trite, this point is a crucial one all the same.

In the second place, it must be recalled that legal norms do not automatically yield the results expected by those who draw them up (or to put it better, draw up the texts from which the norms will emerge). This is particularly true in constitutional matters, since constitutional law is a political law, made by those to whom it applies. Moreover, it is a law of authorisations more than a law of strict obligations (ought is often more of a may). Thus, for example, there is no certainty that members of parliament will make use of the new instruments at their disposal. Experience shows that they lack the will to fully exercise all the opportunities that formal law provides them with. Guy Carcassonne often says, rightly, that the French parliament has powers enough but a shortage of determined members to exercise them. He observes that the French practice of holding multiple offices is a curb on the member investing fully in his or her national mandate. In short, it is not enough to have this full range of powers, it is necessary to have the will to use them. In this respect, only profound changes in behaviour and especially in the preparatory legislative phase, ahead of or in parallel to official procedures (in committee or in public sittings) allowing real consultation between government and the majority, can work any useful effect on the new constitutional provisions. Likewise as regards scrutiny, whether it arises from the majority or the opposition. Time will tell whether the French parliamentary personnel will change on this point.

Then, more fundamentally, the idea of negative parliamentarism, as seen, relates to the articulation of a technical aspect, that of the law of the constitution, which is in this respect essentially static, and of a dynamic aspect, which is that of the system of government, itself produced by the combination of legal rules among themselves and practices and behaviours – sometimes convergent, sometimes disorderly – of the actors in the constitutional theatre. Now, however substantial it may be, the constitutional revision has not affected the first aspect, that of the law of the Constitution. It would be highly self-deluding (a natural inclination of jurists) to consider this change is self-sufficient, at least if one claims – as has to be the case, since this was the purpose of this revision – to evaluate the real effects of the reform on the country’s parliamentary system. Yet, it remains to be appraised how it fits in with the system of government and how it is likely to affect it in any particular direction, whereas it was not thought up (either by the Balladur Committee or by President Sarkozy who initiated it) to completely upset the system. From this perspective, through their very abundance, the formal changes made in 2008–2009 to the situation of the French parliament risk causing disappointment and disillusion should they ultimately fail to bring about marked changes in the balance of the constitutional system.

Now, precisely, while there is no doubt that the legal constraints weighing so far on the French parliamentary assemblies have been considerably loosened, that the negative logic of parliamentarism has sometimes been technically toned down or even reversed, that, subsequently, the chambers have greater leeway both in legislative matters and in scrutiny over the government and administration, and that finally certain changes sometimes benefit the opposition directly, even so the 2008 reform has not truly affected the central armature of the regime, the general equilibrium between the president, the government and the parliament, in other words the presidentialist over-determination of Fifth Republic parliamentarism. By virtue of the general organisation of powers and given French constitutional culture over fifty years, the presidency seems destined to continue to play a major role in determining policy. Besides, all of the main actors in political life are pulling in this direction. That being so, there is every ground to think that the government will continue to be what it has been since de Gaulle: an organ in the service of the President of the Republic. The parliament will not become more involved in its appointment. The fundamental consequence is that the government, even in partisan tune with the parliamentary majority, will continue to see itself as duty-bound to the head of state rather than as the emanation of parliament; it will continue to distrust initiatives from the members of its majority and to seek to control or thwart them. Maintaining the presidential ‘lockdown’ will therefore probably make any real change in spirit difficult and will probably contribute to perpetuating, for the most part, the negative culture of parliamentarism in France. But the life of constitutional institutions has the specificity of being open and it may be that this relatively pessimistic forecast will prove wrong. Which would be no cause for complaint.

[1] ‘Positive law is not law laid down by a legislator, it is the law in force, that is, law as it is applied…’, wrote Capitant (‘Le droit constitutionnel non écrit’, in Mélanges François Gény, t. III (Paris, Sirey, 1934), republished in his Ecrits d’entre-deux-guerres (Paris, Ed. Panthéon-Assas, 2004) p. 298). C. Eisenmann, ‘Droit constitutionnel et science politique’ (1957) RIHPC, pp. 72–85 (republished in his Ecrits de théorie du droit, de droit constitutionnel et d’idées politiques (Paris, Ed. Panthéon-Assas, 2002) pp. 511-524). Eisenmann wrote in response to proponents of a discipline of political science that was gaining its autonomy from the law faculties and who argued for the need to grasp the real world, actual politico-constitutional life.

[2] An empirical observation made by all specialists of parliamentary law since Eugène Pierre and corroborated by the works of Pierre Avril, Jean Gicquel and Jean-Louis Pezant. It squares in certain regards with Michel Troper’s theoretical analyses of the role of volition in the interpretation of law. 

[3] Especially several organic laws and ordinary statutes adopted in 2009 and 2010 as well as alterations to the rules and regulations of the assemblies in May and June 2009.

[4] This term does not in itself imply, to my mind, a disparaging assessment of the type of parliamentarism of the Fifth Republic. It is designed simply to bring out what seems to me, in strictly scientific and descriptive terms (in what a stringent positivist would mean by that) is its principal intrinsic character. That I might be led besides to frame an essential critical or even ‘unfavourable’ overall judgement of this parliamentarism belongs to another register of ‘commitment’ or de lege ferenda.

[5] I hope I will be forgiven a reference to my article ‘Le Prince inapprivoisé. De l’indétermination structurelle de la Présidence de la Ve République’ (2007) 44 Droits, p. 101-137.

[6] For a general presentation see D. Baranger and A. Le Divellec, ‘Régime parlementaire’ in D. Chagnollaud and M. Troper (eds) Traité international de droit constitutionnel, t. II (Paris, Dalloz, 2012, p. 160-193).

[7] There is cause to be more specific and like Philippe Lauvaux (Le parlementarisme (Paris, PUF, 1997) and ‘L’évolution des fonctions exécutives en Europe’ (1994) 268 Cahiers français, p. 30-43) distinguish between classical dualism and renewed dualism and besides between classical monism and pure monism (a range of shadings can separate these broad categories which are merely landmarks, not pure chemical essences).

[8] This final criterion, which is very widely employed in French scholarship, is also the least definite: it is contingent because eminently subject to change.

[9] One and the same system may be a formally dualist parliamentary system, may work according to a monist schema, be weakly or non-rationalised and a majority system (e.g. Austria and Luxembourg). Or it may be in form and in practice monist, rationalised and non-majority (e.g. Hungary).

[10] Jean-Claude Colliard has refined the analysis by proposing to distinguish between the level of structuring and of stability of the party system (Les régimes parlementaires contemporains (Paris, Presses de Sciences Po, 1978) p. 278 ff).

[11] This point is emphasised by R. Fusilier in his major book, Les monarchies parlementaires (Paris, Les éditions ouvrières, 1960) esp. p. 29. But it is also found in more recent analyses such as T. Bergman, ‘Constitutional design and government formation: the expected consequences of negative parliamentarism’ (1993) 16(4) Scandinavian Political Studies, pp. 205-304; Joachim Raschke and Ralf Tils, Politische Strategie : eine Grundlegung (Wiesbaden, Verlag für Sozialwissenschaften, 2007) pp. 189-190. The Report on the role of the opposition in a democratic parliament (CDL-AD (2010) 025) adopted by the European Commission for democracy through law (Venice Commission) at its 84th plenary session in October 2010 also referred to it (Study no. 497/2008, p. 7).

[12] See P. Lauvaux, Parlementarisme rationalisé et stabilité du pouvoir exécutif (Brussels, Bruylant, 1988) p. 103 ff. In Norway, the principle of ministerial accountability established in the late 19th century was long purely conventional. It has recently been enshrined in written law: the new article 15 of the 1814 Constitution (revised 30 March 2007) provides: ‘Any person who holds a seat in the Council of State has the duty to submit his application to resign once the Storting has passed a vote of no confidence against that Member of the Council of State or against the Council of State as a whole. The King is bound to grant such an application to resign. Once the Storting has passed a vote of no confidence, only such business may be conducted as is required for the proper discharge of duties.’

[13] M. Weber, ‘Parliament and government in a reconstructed Germany’ [1918] in Economy and Society (Berkeley, University of California Press, 1978) pp. 1381-1469, Oeuvres politiques, French trans S. Kaufmann (Paris, Albin Michel, 2004) p. 307-455 (esp. p. 344).

[14] A. Le Divellec, Le gouvernement parlementaire en Allemagne (Paris, LGDJ, 2004), pp. 20-35.

[15] ‘Les monarchies : inventaires des types’ (1996) 78 Pouvoirs, pp. 23-41.

[16] This was particularly marked in the written constitutions of 19th-century German monarchies, all of which specified that the material domain of the formal law (i.e. requiring the approval of the houses of parliament) covered matters affecting the law of property and the limitation of civil liberties. For example, the Constitution of the Kingdom of Bavaria of 26 May 1818 (Part VII, art 2) stated ‘Without the council and the agreement of the Estates of the Kingdom, no general law pertaining to personal freedom or the property of the citizens of the State may be enacted, nor may an existing law be changed, authentically elucidated or abolished’.

[17] Although this countervailing power might sometimes seem artificial, insofar as, in important countries (especially Walpole’s England up until the first Reform Act and Italy in the second half of the 19th century) the Crown largely controlled elections and had the capacity to ‘compose’ the chamber.

[18] ‘[T]he first part of the duty of Parliament is the choice of the cabinet who are to administer the affairs of the country’. ‘The Unseen Work of Parliament’, The Economist, 9 February 1861, republished in The Collected Works of Walter Bagehot, vol. VI (London, The Economist, 1974) p. 46.

[19] Thus notably with the system of sessions, which were initially closely controlled by the monarch, who alone had the power to convene them, which, in most republics, have evolved into a right of the chambers to assemble freely. Likewise for the power to appoint their president and their officials themselves.

[20] Sometimes, but more rarely, all of the ministers (case of several German Länder from 1919 and of the Austrian Federal Constitution from 1920 to 1929).

[21] This is in some countries further consolidated by a constructive motion of censure type of mechanism invented by the Germans after 1945, legally guaranteeing the simultaneity of the election of a new prime minister and the overthrow of a government, theoretically compelling the parliament to face up to its responsibilities.

[22] More or less successfully. It has happened that this system is powerless to produce the hoped-for effects, especially when the parliamentary forces prove unable to reach understandings.

[23] Sometimes it is the constitutional court that requires assemblies not to forego their own constitutionally guaranteed competence. This was the case of the German Federal Constitution on the ‘essentiality theory’ (Wesentlichkeitstheorie) or of the French Constitutional Council on ‘the legislator’s negative incompetence’.

[24] Thus, for example, the 1849 Danish Constitution (in its 1866 version) (§18) empowered the King alone to declare war and make peace, conclude alliances and trade agreements (the approval of the then bicameral parliament only being necessary for the ceding of territory and undertakings modifying the existing state of constitutional law). In 1920, the Constitution was modified to subordinate the exercise of all those powers to the assent of the chambers. Similarly in the Netherlands where the principle of parliamentary approbation prior to ratification of treaties entered into by the King was established systematically in 1922 (§58 of the 1815 Constitution consolidated in 1887).

[25] Thus, for example, the abundant case law of the German Constitutional Court since its ruling of 20 July 1994, enshrined in federal legislation, on the participation of parliament in the decision to commit armed forces abroad (Parlamentsbeteiligungsgesetz) of 18 March 2005.

[26] A position defended emphatically especially in its ruling on the Maastricht Treaty (BVerfGE 89, 155 of 12 October 1993) and then on the accompanying statute of the Lisbon Treaty (BVerfGE of 30 June 2009) in which the Court forced the federal legislature to revise its legislation on the reinforcement of the rights of the Bundestag and the Bundesrat (finally promulgated on 22 September 2009). More recently, it has asserted a Bundestag reserve over the European Financial Stability Fund (decision of 7 September 2011, VerfG, 2 BvR987/10).

[27] Thus especially for the designation of constitutional judges by parliament, either wholly (e.g. Austria from 1920 to 1929, FRG, Poland) or in part (Italy, Austria since 1929, Spain, Portugal).

[28] Thus it is for the ‘backstage’ relations between government and members of parliament, especially those of the majority, to favour dialogue and prepare legislative or other decisions ahead of time.

[29] In this respect this anchoring concerns legislation as much as scrutiny (insofar as these two functions of parliaments can been clearly told apart), affecting both the majority, whose vocation is to govern alongside the cabinet, and the opposition whose mission is to force the government and its majority to explain their decisions.

[30] Even though there have been nuances in the way this break off is presented: whereas Michel Debré highlighted the idea that through the draft for the new Constitution ‘the government wants to renovate the parliamentary regime’ (speech before the State Council, 27 August 1958, reproduced in Documents pour servir à l’histoire de l’élaboration de la Constitution du 4 octobre 1958, vol. III (Paris, La Documentation Française, 1991) p. 255, General de Gaulle never spoke of a parliamentary regime, other than to claim (in December 1962) that ‘the parliamentary regime is over’ (A. Peyrefitte, C’était de Gaulle, t. I (Paris, Fayard, 1994) p. 274).

[31] ‘The preliminary draft submitted to you is designed to fine-tune a sanitised parliamentary regime. It is a parliamentary regime since both the principle of government being accountable to parliament, which is included in the constitutional law, is the very definition of the parliamentary regime. But it must be a sanitised regime in that it seems desirable to avoid governmental instability and, for that, to give the Government the means to govern’ (Session of 31 July 1958 of the Consultative Constitutional Committee, reproduced in Documents pour servir à l’histoire de l’élaboration de la Constitution du 4 octobre 1958, vol. II (Paris, La Documentation Française, 1988) p. 67.

[32] Le régime parlementaire dans les constitutions européennes d’après-guerre (Paris, Sirey, 1937).

[33] Additional sessions may be held but may be refused at the discretion of the President of the Republic (according to General de Gaulle’s interpretation in 1960).

[34] Order no. 58-1100 of 17 November 1958, art. 6.

[35] Article 42 C (1), in providing that ‘the discussion of bills before the first assembly seized bears on the text presented by the Government’ is another fine illustration of negative logic. Even if the committee vigorously wishes to change the government text, it is for members of parliament in plenary session to show their will to impose their views by voting.

[36] See C. Vintzel, Les armes du gouvernement dans la procédure législative. Etude comparée : Allemagne, France, Italie, Royaume-Uni (Paris, Dalloz, 2011).

[37] By virtue of the judicial decision of the Constitutional Council of 17, 18 and 24 June 1959 (no. 59-2 DC).

[38] On this process, see below.

[39] One of de Gaulle’s asserted leitmotivs, from his 1946 Bayeux Speech onwards, but that remained ambiguous about how it should be reconciled with keeping the government accountable to parliament.

[40] Except for the specific case of cohabitation, a configuration where the actors accept to suspend presidential power to freely choose the prime minister so as to comply with the ballot-box verdict of elections to the National Assembly. It is then accepted that the choice remains, but is narrowly circumscribed to someone from the new majority, and in general to whoever is implicitly designated by that majority to the head of state (Mr Chirac in 1986, Mr Balladur in 1993 and Mr Jospin in 1997).

[41] Albeit hesitantly by Michel Debré. Conversely, Georges Pompidou showed himself to be immediately won over by de Gaulle’s interpretation, as was M. Couve de Murville in 1968.

[42] When Pompidou was called on to form a third government in January 1966 (after de Gaulle was re-elected president), parliament was not sitting and was not convened. At the opening of the ordinary session at the start of April, he refrained from asking for a vote of confidence (contrary to what he had done in April and then November 1962), claiming that the ‘letter and spirit of the 1958 Constitution intended that the Government should be entirely free’ to ask for a vote or not (National Assembly, 13 April 1966, reproduced in D. Maus, Les grands textes de la pratique institutionelle de la Ve République, 5th edn (Paris, La Documentation française, 1990) Doc. 49-101, p. 166.) The opposition filed a motion of censure (focused on condemnation of the President of the Republic’s decision to withdraw French forces from integrated NATO command) on the basis of article 49 (2), which obtained only 137 votes (out of the 242 required).

[43] P. Avril, Les conventions de la constitution (Paris, PUF, 1997) p. 115.

[44] It should be said that the government, from the second legislature onwards, endeavoured to limit its own troops from stepping out of line, if need be by wielding legal instruments such as, in legislative matters, single in toto voting, government orders or even article 49 (3).

[45] This conception is essentially ‘psychological’. Elites have interiorised presidentialism to such an extent that they almost overlook the value of law or constitutional structure and finally hold parliament’s role to be a negligible quantity. We are operating here on the register of mind-sets, nurtured by near ideological discourse. It is so much more surprising (or remarkable), in this respect, that the Prime Minister, Mr François Fillon, was recently able to say before the majority group that his legitimacy depended on the President but also on the majority (Le Figaro 24 March 2010). But such declarations are extremely rare in the Fifth Republic.

[46] P. Avril, ‘La majorité parlementaire ?’ (1994) 68 Pouvoirs p. 46. The eminent author goes on ‘it is less governing than governed’ and ‘instrumentalised’.

[47] An innovation compounding the implicit constitutionalisation of sessions of ‘questions to the government’ introduced by agreement in 1974 (see P. Avril and J. Gicquel, Droit parlementaire, 4th edn (Paris, Montchrestien, 2010), p. 299.

[48] Statute 72-553 of 3 July 1972 creating ‘consultative parliamentary delegation for French radio and television broadcasting’ (transformation of the parliamentary representation previously provided for by the statute of 27 June 1964). See Avril and Gicquel, Droit parlementaire, p. 114.

[49] The primary examples being the specialised delegation in European affairs (since 1979), the parliamentary office for evaluation of scientific and technological choices (since 1983), the delegation for women’s rights and equal opportunities (since 1999) and more recently the parliamentary delegation on the intelligence services (since 2007).

[50] See J.-E. Gicquel, ‘Vers la fin de la maîtrise gouvernementale en matière de fixation de l’ordre du jour des assemblées parlementaires’ (8 July 1999) 135 Les Petites Affiches, pp. 12-20.

[51] The letter of assignment from the President of the Republic of 18 July 2007 set the deadline for the committee to file its conclusions and proposals for 1 November of that same year.

[52] For a systematic review, see J.-P. Camby, P. Fraisseix and J. Gicquel (eds), La révision de 2008 : Une nouvelle Constitution ? (Paris, L.G.D.J., 2011).

[53] Several texts, including the formal Constitution (arts 56, 65 and 71-1 C) but also organic laws (in particular organic law no 2010-837 of 23 July 2010) or ordinary statutes set out the list of positions in question.

[54] It will be noted incidentally that unlike the rules in the Senate, those in the National Assembly specify that, exceptions apart, a public interview of the candidate must be arranged (art. 29-1 RAN).

[55] See P. Wachsmann, ‘Sur la composition du Conseil constitutionnel’ (2011) vol. III Jus Politicum, p. 103–136 (and electronic version in no. 5, 2011).

[56] The constitutional court has not been called on to decide on questions of the kind and it is probably better that way because it would have become mired in an irreducibly political controversy.

[57] Both assemblies thus approved by a large majority on 12 July 2011 the continued engagement of French troops in Libya begun in March 2011 under a UN Security Council resolution.

[58] Statute of 18 March 2005, passed after abundant case law of the Federal Constitutional Court since 1994. Similar solutions hold especially in Denmark, Sweden, Ireland and Austria.

[59] The minority position in the Senate is less favourable since if the ‘right to create a commission of enquiry or a mission of information per parliamentary year’ is asserted (art. 6 bis RS), it still requires a draft resolution for this to be passed (art. 11 (1) RS) with the result that the majority retains control over the decision.

[60] Art. 139 (2) and (3) RAN if legal proceedings are underway as to the actions that caused the proposal to be filed. This principle is posited by article 6 (3) of Order no. 58-1100 of 17 November 1958 as amended.

[61] Which happened on 26 November 2009 over the application by the socialist group for a commission of enquiry into opinion polls commissioned by the President of the Republic.

[62] As was the case in January 2010, again over the question of the Elysée’s opinion polls, so that the opposition withdrew its application.

[63] Art. 132 RAN and art. 39 (3 bis) RS.

[64] This article was introduced in the National Assembly which wanted to offset its refusal to accept resolutions of a general character. It was maintained even though resolutions were finally included in the reform.

[65] For constitutional, financial or social security financing bills.

[66] A. Le Divellec, ‘Le gouvernement, portion dirigeante du Parlement. Quelques aspects de la réception juridique hésitante du modèle de Westminster dans les Etats européens (2009) 1 Jus Politicum, p. 185-225 (esp. 211-214).