http://journals.cambridge.org Downloaded: 26 Nov 2010 IP address: 128.122.149.54
223Concrete Constitutional Review in France
Concrete Review as Indirect Constitutional
Complaint in French Constitutional Law:
A Comparative Perspective
Otto Pfersmann*
1958: a priori review of parliamentary legislation to keep Parliament within its limited
legislative competences – 2008: a posteriori review to protect citizens’ rights
and freedoms – Varieties of concrete review – Originality of the new French procedure
– An indirect constitutional complaint: raised by a litigating party, transmitted
by an ordinary court and decided by the Constitutional Council – Obstacle to transmission:
declaration of conformity in the motives of an earlier decision by Constitutional
Council – Conceptual, logical and legal reasons for a strict distinction
between normative content and justificatory discourse in judgments – Setting on
par of motives and operative part transforms French legal system into a specific
form of common law and weakens the Rule of Law
France is often said to stay outside many common legal evolutions. This applied
especially to constitutional review of formal legislation. It was introduced as relatively
late as 1958, and then only abstract and a priori, thus in a very different mode
than in most other European systems. In 2008, France adopted an extended constitutional
reform, finally introducing review of formal parliamentary legislation a
posteriori with regard to constitutional ‘rights and freedoms’. Some will say that this
new procedure brings the French legal order closer to European normality, while
others may claim that it is again sui generis, having no real equivalent in other countries
or international structures. In fact, the French system is very well suited to an
extended comparison. However, some specific elements may not be easy to capture
and require special analysis.
Other elements can, at first sight, be rather easily stated. On the one hand, the
traditional French ideology of the ‘law’ as an expression of the general will induced
a long-standing reservation against judicial review of parliamentary statEuropean
Constitutional Law Review, 6: 223–248, 2010
© 2010 T.M.C.ASSER PRESS and Contributors doi:10.1017/S1574019610200044
* Professor, University Paris I Panthéon Sorbonne.
http://journals.cambridge.org Downloaded: 26 Nov 2010 IP address: 128.122.149.54
224 Otto Pfersmann EuConst 6 (2010)
utes.1
However, on the other hand, distrust in Parliament’s ability to sustain a
stable government and to provide for coherent legislation prompted de Gaulle to
introduce constitutional review.2
It was not introduced to protect rights and liberties
or a distribution of competences between a centre and decentralised entities.
Instead, it was conceived of as a strictly preventive check on legislation, to prevent
Parliament from overstepping its limited competences (Article 34 Constitution).
It left promulgated statutes completely outside the reach of any judicial challenge.
Referrals could be filed only by the highest political authorities (the President of
the Republic, the Prime Minister and the presidents of the chambers of Parliament),
not by the addressees of the norm. With a very few exceptions,3
the Constitution
provided no ‘rights’ – thus rights could not be invoked as a term of
constitutional reference.4
As is well-known, it was the Constitutional Council (Conseil constitutionnel ) that
introduced rights into the formal Constitution through a revolution in 1971, by
using the Preamble as a norm of reference.5
It should be noted that this already is
1
French constitutional ideology mainly has been shaped by the ‘revolutionary tradition’ under
the Third Republic, which is best expressed by Raymond Carré de Malberg in his La loi, expression de
la volonté générale. Etude sur le concept de la loi dans la Constitution de 1875 (Paris, Sirey 1931; reprint Paris
Economica 1984) and his Confrontation de la théorie de la formation du droit par degré avec les institutions et les
idées consacrées par le droit positif français relativement à sa formation (Paris, Sirey 1933). It contains the
following propositions: 1) ‘law is the expression of the general will’ (Rousseau in the Contrat Social as
well as Art. 6 of Declaration of the Rights of Man and of the Citizen of 1789); 2) Parliament is the
organ which states the general will; 3) all other organs, hence all courts, have to apply the law.
2 The ideology of parliamentary sovereignty coexists paradoxically with the belief that Parliament
is unable to provide governmental stability. The Constitution of 1958 is designed in order to
give the government the means to frame the proceedings of the Parliament. Constitutional review
as intended in this design aims at guaranteeing that the legislature would not overstep its limited
competences (Art. 34). Cf. Louis Favoreu, Didier Maus, Jean-Luc Parodi (eds.), L’écriture de la Constitution
de 1958 (Paris, Economica 1992); Louis Favoreu (ed.), Le domaine de la loi et du règlement (Paris,
Economica 1981).
3 E.g., Art. 1 guarantees the equality of citizens and religious freedom, Art. 3 the right to vote
and Art. 67 entrusts personal liberty to the judiciary.
4 That the Council was not conceived as a court and the Preamble not as a Charter of rights
which could be invoked by the Council, appears clearly in the minutes of the elaboration of constitution:
Comité national chargé de la publication des travaux préparatoires des institutions de la
Ve République, Didier Maus (dir.), Documents pour servir à l’histoire de l’élaboration de la constitution du
4 octobre 1958, 4 vols. (Paris, La Documentation française 1987 squ).
5 In its famous decision 44 DC of 16 July 1971, the Constitutional Council stated that the
Preamble was part of the Constitution and hence that all the rules and principles mentioned or
referred to in it were constitutional exigencies. This not only transformed the Declaration of 1789
and the Preamble of the Constitution of 1946 into formally constitutional documents, but also gave
constitutional status to the ‘Fundamental principles of republican legislation’ (hence the law on the
‘contract of association’ of 1905 became a constitutional guarantee of the liberty of association).
The Preamble, though, was not framed as a part of the normative Constitution; if it were otherwise,
http://journals.cambridge.org Downloaded: 26 Nov 2010 IP address: 128.122.149.54
225Concrete Constitutional Review in France
a controversial statement, as a majority of scholars maintains that the Council
simply ‘consecrated’ what was already there since the very beginning.6
In 1974,
the members of Parliament obtained the right to refer laws to the Council with 60
deputies or 60 senators. This shifted the equilibrium of the Constitution of 1958/
62 in an unexpected and indeed very specific way.
When exercising a priori review, the Council examines an entire bill after its
adoption by both chambers of Parliament within one month. Contrary to the
European standard, a decision of the French constitutional court extends to an
entire statute and provides it with a label of constitutionality, except for those
provisions which are annulled as unconstitutional and for those subjected to a socalled
‘interpretation in conformity’.7
Since 1975, the French system has been unexpectedly enriched by a parallel
system of review of formal legislation. The French Constitution states in Article
55 that
it would not have been labelled a Preamble, but a catalogue of rights, and it certainly would have been
more cautiously and less generously drafted. The introduction of these elements into the Constitution
was not operated according to the rules governing the modification of the Constitution, and
yet these elements are since then considered to be part of the Constitution. Thus the Constitutional
Council modified the Constitution without having been empowered to do so. This was thus a revolution
in the legal meaning of the term (Cf. Otto Pfersmann, in Louis Favoreu (ed.), Droit constitutionnel
(Paris, Dalloz 1998, 1st edn.; 2008, 11th edn.).
6 The majority opinion of today had to fight for the recognition of constitutional review as a
mechanism for enforcing the Constitution. It is understandable that it presented the Preamble as
part of the Constitution since the beginning. Cf. Louis Favoreu, Loïc Philip, Les grandes décisions du
Conseil constitutionnel (Paris, Dalloz 1975, 1st edn.); in this and all subsequent editions, the relevant
passages can be found under the title ‘Liberté d’association’, for instance in the edition of 2009, p.
180-199.
7 The so-called interpretation in conformity (Verfassungskonforme Interpretation, decision de conformité sous
reserve, sentenza interpretativa) consists in the ruling that a legislative provision which has at least two
different meanings is considered to be constitutional under one (or more) meaning(s) and not under
another meaning (or meanings). This today seems to be a common European practice. It is practised
by all constitutional jurisdictions and by some of them in a highly extensive and differentiated
way, like in Italy (cf. e.g., Thierry di Manno, Le juge constitutionnel et la technique des de´cisions ‘interpre´tatives’
en France et en Italie (Paris, Economica 1997); Riccardo Guastini, Il diritto come linguaggio (Torino,
Giappichelli 2006); idem., ‘Teoria e ideologia dell’interpretazione costituzionale’, in Giuisprudenza
costituzionale (2006), p. 743 et seq.; Giusi Sorrenti, L’interpretazione conforme a costituzione (Milano, Giuffrè
2006). This technique supposes indeed interpretation in order to identify the various meanings of the
relevant provisions, but its very operation consists in a partial annulment and in certain cases the
substitution by alternative provisions. It hence is not review, but alternative law-making. It requires
thus a constitutional authorisation, which does not exist, i.e., nowhere does any constitution or
treaty empower courts to ‘decide which of the norms expressed by a legislative provision may be
upheld, which of the norms so expressed may be annulled and which of the norms so expressed
may be replaced by others, enacted by the court itself.’ In this respect, France shares the European
standards: the Constitutional Council, an organ which duty and mission it is to check the respect of
constitutional exigencies, oversteps its constitutional competencies.
http://journals.cambridge.org Downloaded: 26 Nov 2010 IP address: 128.122.149.54
226 Otto Pfersmann EuConst 6 (2010)
Treaties or agreements duly ratified or approved shall, upon publication, prevail
over Acts of Parliament, subject, with respect to each agreement or treaty, to its
application by the other party.
According to the Council this does not mean that it itself has jurisdiction to make
treaties prevail over statutes: circumstances may change and, accordingly, superiority
is subject to modifications, whereas the Council states in a final manner
whether a piece of legislation is constitutional or not.8
The review of conventionality
of statutory provisions has therefore been appropriated by ordinary courts,
which, at least since 1989, do indeed let international treaties prevail over formal
legislation, even if the legislation is subsequent to the treaty.9
On this basis, the
ordinary courts developed a rights-based jurisprudence, especially in reference to
the ECHR. Although strictly opposed to any form of judicial review of promulgated
legislation with respect to the Constitution, French law developed therefore a
system of review of promulgated legislation with respect to international treaties, largely
equivalent to the guarantees offered by the constitutional revolution of 1971.
It was only in 1990, and again in 1993, that attempts were made to introduce
constitutional review a posteriori. Although the proposals were framed in a very
restrictive way, the Senate opposed the initiatives adopted by the National Assembly.
In 2007, the new President of the Republic initiated a large-scale constitutional
reform, mainly intended to bring the legal framework closer to the American
model of presidential government. The idea consisted in subjecting the Prime
Minister to the President, the introduction of hearings of nominees for certain
important positions and of the possibility for the President to address the Assem-
blies.10
To prepare the reform, he formed a committee to report on the
‘modernisation of the institutions’.11
In its report,12
the committee voiced the
need of introducing review a posteriori to repair the consequences of the 1975
decision and to bring ‘hierarchical superiority back to the Constitution’. This was
not part of the original presidential plan, but finally it was accepted by all political
actors and inserted into the formal Constitution.13
The new Article 61-1 reads:
8 Decision 15 Jan. 1975. For the majority reading, cf. Louis Favoreu, Loïc Philip, Les grandes
decisions du Conseil constitutionnel (Dalloz 2009) p. 180-199.
9 Council of State, Nicolo, 20 Oct. 1989.
10 Cf. Otto Pfersmann, ‘Verfassungsrevision in Frankreich’, in Michael Thaler, Harald Stozlechner
(eds.), Verfassungsrevision. Überlegungen zu aktuellen Reformbestrebungen (Vienna, Jan Sramek Verlag 2008)
p. 27-51, with further references.
11 See, concerning the alleged reasons for its creation: .
12 This report can be found at .
13 The whole reform was adopted by Parliament on 21 July 2008, promulgated on 23 and published
on 24 July in the Official Gazette. The constitutional reform act enters into force on 1 of March
2009 or at a later date, according to the relevant provisions. This is case for the new rules concerning
http://journals.cambridge.org Downloaded: 26 Nov 2010 IP address: 128.122.149.54
227Concrete Constitutional Review in France
If, during proceedings in progress before a court of law, it is claimed that a legislative
provision infringes the rights and freedoms guaranteed by the Constitution,
the matter may be referred by the Council of State or by the Court of Cassation to
the Constitutional Council which shall rule within a determined period.
An organic Act shall determine the conditions for the application of the present
Article.
This provision is completed by a new version of Article 62(2), which now states:
A provision declared unconstitutional on the basis of Article 61-1 shall be repealed
as of the publication of the said decision of the Constitutional Council or
as of a subsequent date determined by said decision. The Constitutional Council
shall determine the conditions and the limits according to which the effects produced
by the provision shall be liable to challenge.
These provisions, however, were not to take effect before the entry into force of
an ‘organic act’.14
This act, adopted by Parliament in November 2009, was subject,
as all organic legislation, to compulsory review by the Constitutional Council,
which delivered its decision on 3 December 2009.15
According to the provisions
of the organic act, the reform entered into force on 1 March 2010.
Understanding the new procedure requires a presentation first of the varieties
of concrete review and then of the specificities of the French procedure. In a
third part, I shall revert to one element of the new structure, which until now has
attracted little attention and which formalises a ‘French system of constitutional
constitutional justice, which take effect with the entry into force of the organic law, which concretises
the constitutional setting. Concerning the reform as a whole, see, e.g., Revue Française de Droit Constitutionnel,
‘Après le Comité Balladur – Réviser la Consitution en 2008?’, Numéro hors-série 2008;
Petites Affiches, ‘Une nouvelle Constitution? Commentaire article par article du texte de la loi du 23
juillet 2008 de modernisation des institutions de la Ve République’, 19 Dec. 2008 ; Revue Française de
Droit Constitutionnel 82 (2010), ‘Une nouvelle Vème République’, in particular: Xavier Philippe, ‘La
question prioritaire de constitutionnalité…Réflexions après l’adoption de la loi organique’, p. 273-
287.
14 In the French legal system organic laws are statutes adopted according to a special procedure
framed by Art. 46 of the Constitution. At least fifteen days have to pass between the introduction
of the bill and its debating and voting; moreover, if, after the usual vote by the two chambers, the
National Assembly is given the final say; it can only accept the bill with an absolute majority of its
members. They cannot be promulgated until the Constitutional Council has declared they are in
conformity with the Constitution.
15 Constitutional Council, decision n° 2009 – 595 DC. The whole documentary file is accessible
under: . The organic law (‘Loi organique no 2009-1523 du 10 décembre 2009 relative à l’application
de l’article 61-1 de la Constitution’) and the decision are published in the Official Gazette on
11 Dec. 2009, accessible under: .
http://journals.cambridge.org Downloaded: 26 Nov 2010 IP address: 128.122.149.54
228 Otto Pfersmann EuConst 6 (2010)
common law’, or so I shall contend. I shall conclude with some general remarks
on the possible consequences of the reform.
Comparative settings: Parties in concrete review
Concrete review is the examination of the constitutionality of primary legislative
provisions in a concrete case by an organ with judicial status. The distinctive features
of this kind of procedure are therefore: 1) litigation in a court concerning a
given legal question, 2) primary legislation relevant to the case, in respect to which
the court has doubts concerning its conformity with the formal constitution, 3) a
judicial decision stating, possibly among other things, whether the possibly unconstitutional
provision ought to be applied or not. This general structure leaves
room for many varieties. The issue is always decided by a court, but this may be
the same court as the one which has to settle the case at hand, or another one
specially entrusted with the constitutional question. The unconstitutional provision
may be declared inapplicable in the concrete case or annulled and thus eliminated
from the legal system.
The constitutional question itself may be raised by the court settling the original
litigation (judex a quo), by the parties or by both of them. In most European
systems however, the judex a quo itself raises the question and decides whether or
not the issue should be referred to another, specialised – constitutional – court.
The first court, and not the parties in the original litigation, then becomes a party
in a separate procedure. In the new French procedure, however, the original parties
are also the parties in the separate constitutional procedure. Nevertheless, it is
the judex a quo who decides whether or not the so-called ‘priority constitutional
question’ (question prioritaire de constitutionnalité or QPC)16
has to be transmitted to
the Constitutional Council. One can thus distinguish two traditional settings of
concrete review, which may be termed immediate concrete review and preliminary ruling:
courts as parties. To these may be added a new type of concrete review, which is in
fact an indirect constitutional complaint, of which the new French procedure is a spe-
cies.
Immediate concrete review
I call immediate concrete review the procedure in which (one of) the parties in the
concrete case can, as such, request the review of relevant statutory and allegedly
unconstitutional provisions. Mostly, but not necessarily, a check will be requested
against fundamental rights. In this kind of review, the constitutional question is
16 This terminology has been introduced in the Organic Act in order to show that constitutionality
has to be decided prior to conventionality.
http://journals.cambridge.org Downloaded: 26 Nov 2010 IP address: 128.122.149.54
229Concrete Constitutional Review in France
not separated from the litigation at stake, but has to be resolved in the course of
the solution of the case and the settling of the question of constitutionality is a
premise for other elements of the ruling. The constitutional question is not procedurally
but legally prior to the questions on which the constitutional norm has
an impact. Once the possibility of raising constitutional questions is open to the
litigating parties and once the court has the power to exert constitutional review,
there is no need for special procedures (although of course this is not theoretically
excluded). In short, immediate concrete review is settling litigation while directly
minding constitutional requirements. The problem with general immediate concrete
review is not procedural and hence does not raise questions as to the status
of the parties. As is well-known, the issues are different and they do not only
concern concrete review, but more generally the legal status of a declaration of
unconstitutionality and hence the legal destiny of the norm identified as unconstitutional.
Although the judge finally sets the issue of the case at stake, he does not
finally determine the legal destiny of the unconstitutional norm.17
As there is only one procedure, immediate concrete review raises no questions
as to the status of the parties in the constitutional litigation. This is different in the
other varieties of concrete review.
Objective concrete review: courts as parties
The distinctive feature of constitutional review in the European model of constitutional
justice18
is, or at least was initially, the fact that the legal destiny of an
17 The structural problem of the indeterminacy of the legal destiny of the unconstitutional
norm is particularly apparent in the US, although this does not seem to trouble American scholarship
in any particular way. The link between the fate of unconstitutional provisions and immediate
concrete review is not, however, an intrinsic element of the ‘American’ model: it also can be found
in Belgium, which has centralised review, and a system with a general judicial review-competence
could theoretically allow courts to definitely eliminate unconstitutional provisions. Cf. Otto
Pfersmann, ‘Modèles organocentriques et modèles normocentriques de la justice constitutionnelle
en droit comparé’, in Mélanges Francis Delpérée (Brussels, Bruylant 2007), p. 1131-1145.
18 The concept of the ‘European’ or ‘Austrian model of constitutional justice (or: review)’ is
usually defined by the following elements: a special court, placed outside the ordinary judiciary, with
exclusive competence in matters related to the application of the formal Constitution and especially
the power to annul unconstitutional statutory provisions erga omnes, whereas the American model
would be characterised by the fact that all judicial organs have constitutional competence and that
their decisions are valid only inter partes (a classic reference may be Mauro Cappelletti, Judicial Review
in the Contemporary World (Indianapolis, Bobbs-Merrill Company 1971)). This classification is problematic
as it mixes heterogeneous elements, which are, besides this, much too large and undetermined
(cf. Pfersmann, supra n. 17). In the French debate, particular attention has been paid to
showing that the Constitutional Council truly belonged to the European model. The discussion
concerned not only the question of comparative classification, but of the desirable evolution of
constitutional review in France. Louis Favoreu has certainly been the most influential writer in this
http://journals.cambridge.org Downloaded: 26 Nov 2010 IP address: 128.122.149.54
230 Otto Pfersmann EuConst 6 (2010)
unconstitutional provision is definitively settled in a separate procedure.19
In this
case, the constitutional question cannot be integrated into one and the same ruling
because the issue of the termination of validity requires a separate decision.
Theoretically this competence could have been entrusted to ordinary courts, but
the second element of distinctiveness of the European model consists in a distribution
of competences between on the one hand ordinary courts and on the
other hand one constitutional court with a special and exclusive constitutional
jurisdiction. This raises the question of who will be a party in the separate constitutional
procedure.
Concrete constitutional review is subjective, when the constitutional procedure
is driven by a complainant who considers that his constitutional position, generally
his fundamental rights, is violated; it is objective if the referral is not linked to
the individual interest of a particular person.
In most systems, concrete review is objective. An ordinary court (a quo) identifies
a possibly unconstitutional provision, which has an impact on the outcome of
the concrete case. The whole constitution is a possible reference for review. The
court then has the authorisation or possibly the duty to transmit, as a party, the
question or the request to the constitutional court in a new and separate procedure.
The ordinary court has to check the conditions of admissibility and to sustain
the argument. The constitutional court will again check the admissibility and
rule on the request filed by the first court as a party. In such cases, the original
litigants are possibly granted rights as secondary parties. They may present briefs and
may be present in the audience, if the constitutional court has public hearings.
debate (cf. ‘Symposium In honour of the late Louis Favoreu: France’s exceptionalism in constitutional
review’, International Journal of Constitutional Law 2007 5(1)), cf. e.g., his ‘Rapport général
introductif’, Revue internationale de droit comparé. Vol. 33 (1981), p. 255-281 (for alternative accounts:
Olivier Jouanjan, ‘Modèles et représentations de la justice constitutionnelle en France : un bilan
critique’, Jus Politicum, n°2, 2009, p. 1-25; Dominique Rousseau, ‘The Conseil Constitutionnel confronted
with comparative law and the theory of constitutional justice (or Louis Favoreu’s untenable
paradoxes’, ICON, 5 (2007), p. 28-43). Favoreu claimed that because French review was suited to
the European model it should not be changed, i.e., it should not be opened to concrete review (see
his contribution in ‘L’exception d’inconstitutionnalité (Le projet de réforme de la saisine du Conseil
constitutionnel’, Revue française de droit constitutionnel, 1990, no. 4, p. 581-617 as well as his ‘Sur
l’introduction hypothétique du recours individuel devant le Conseil constitutionnel, Cahiers du Conseil
constitutionnel no. 10 (2001), p. 163-169.
19 The point is not only, contrary to a widespread opinion concerning the nature of the ‘European
model’, that the ‘declaration of unconstitutionality’ be centralised, but that the unconstitutional
provision be definitely removed from the legal system, both as a norm and as the text which
expresses the norm. This constitutive element of the Austrian model has progressively been weakened
through the techniques of ‘interpretation in conformity with the constitution’ (cf. supra n. 7)
and through the modification of the nature of the judicial opinion becoming a normative statement
(without constitutional empowerment; cf. the third part of this article).
http://journals.cambridge.org Downloaded: 26 Nov 2010 IP address: 128.122.149.54
231Concrete Constitutional Review in France
In certain systems, objective concrete review leads to paradoxical features. For
instance in Austria20
the Constitutional Court may, as a first court, file ‘ex officio’ a
request to itself as Constitutional Court, if it considers that a statutory provision
it has to apply in a case pending before it may possibly be unconstitutional (Article
140(1) Austrian Constitution). Ex officio here means that the Court first has to
suspend the case at hand and to specify the doubts concerning the constitutionality
of the statutory provision concerned. It then acts as a party to a new procedure
in which the same Court, but now in another capacity, checks whether the ‘decision
to interrupt the original procedure’ meets the requirements of admissibility.
If so, the arguments of the Court as judge a quo will be examined and the constitutional
question decided in substance by the Court as judge ad quem, as the Constitutional
Court proper.
In certain cases, the litigating parties may ask the court to lodge a request with
the constitutional court. This is the case in Italy and Spain,21
where provisions can
be checked against the whole Constitution. Nevertheless, although the capacity to
raise the issue of constitutionality pertains both to the litigating parties and the
court which has to settle the original case, only the latter can refer the constitutional
question to the constitutional court. These countries therefore also have
objective constitutional review procedures, although, as in Germany and Austria,22
the
litigating parties are provided with derivative rights as secondary parties.
Subjective concrete review: indirect constitutional complaints
The French procedure certainly is a variety of concrete constitutional review according
to the European model. The original litigation is separated from the litigation
before the Constitutional Council, which rules exclusively on the
constitutionality of a contested provision. The procedure starts before an ordinary
court (administrative, civil or penal), which under certain circumstances may
transmit the question to the highest tribunals to which it pertains. The Council of
State, i.e., the highest administrative court, or the Court of Cassation, i.e., the
20 On the Austrian case, cf. Otto Pfersmann, ‘Le contrôle concret des normes législatives en
Autriche’, in Xavier Philippe (ed.), Le contrôle de constitutionnalité par voie préjudicielle (Paris, PUAM
2009), p. 91-107.
21 Italy: Art. 1 Constitutional Law no. 1 from 9 Feb. 1948; Spain: Art. 35 Organic law 2/1979
from 3 Oct.
22 Germany, Art 82(3) of the Law on the Federal Constitutional Court: ‘The Federal Constitutional
Court grants the participants in the litigation before the tribunal, which has filed the referral,
the opportunity to present their views, it invites them to the oral hearing and gives the council,
present at the proceedings, the floor.’ Austria, Art. 63(1) of the Federal law on the Constitutional
Court: ‘The President calls for the hearing. The Federal government as defendant of the challenged
legislation and the applicant are invited. (…) If the referral is introduced by a tribunal (…), the
parties to the litigation have also to be invited.’
http://journals.cambridge.org Downloaded: 26 Nov 2010 IP address: 128.122.149.54
232 Otto Pfersmann EuConst 6 (2010)
highest civil and penal court, as the case may be, transmits the file to the Constitutional
Council if admissibility conditions are fulfilled. Yet, although courts refer
the question to the constitutional judge, this is not a case of objective review.
The French procedure differs in an essential way from comparable procedures
in Spain and even Italy, where the Constitution explicitly gives the parties the right
to raise the constitutional issue before the ordinary courts.23
Whereas in Spain
and Italy the competence to raise the question is shared by the ordinary courts and
the parties to the original litigation, this right in France is the exclusive competence
of the parties; yet only the courts are allowed (and even obliged) to transmit
the question under specific circumstances. In short, the original parties are also
the primary parties in the constitutional litigation, even though they are not themselves
empowered to transmit the referral (‘question’). Raising and transmitting here
are entirely split competences.
Specific ideological reasons have led to this construction, which I propose to
term ‘indirect constitutional complaint’. In the French constitutional debate, objective
concrete constitutional review has never been a real issue. Neither politics
nor constitutional scholarship was interested in giving ordinary courts the initiative
of raising constitutional questions before the Council. The ordinary courts
are not considered as institutions striving at promoting constitutional justice. This
is probably due to a contradictory perception of the ‘judge’, which in turn reflects
a contradictory idea of the Constitutional rule of law.24
23
Cf. supra n. 21.
24
‘Rule of law’ is used as a theoretical concept. It means a legal order characterised by the
following elements: 1) a high degree of normative determinacy; 2) a clearly articulated ‘calculus of
defaults’ (norms stating the legal consequences of deficiencies in other norms: annulment, modification
or non-application); 3) review of normative concretisation, i.e., of valid norms lacking conformity
(presenting some deficiency with respect to higher norms); 4) the exigency to justify applicative
(or, more technically, concretising) norm-production (cf. Otto Pfersmann, ‘Prolégomènes pour une
théorie normativiste de «l’Etat de droit»’, in: Olivier Jouanjan (ed.), Figures de l’Etat de droit. Le
Rechtsstaat dans l’histoire intellectuelle et constitutionnelle de l’Allemagne (Presses Universitaires de Strasbourg
2001) p. 53-78). The Constitutional rule of law integrates the application of the formal constitution
into the rule of law. The rule of law is an ideal which is impossible to realise, because a comprehensive
review of normative concretisation must always end somewhere with a decision concerning
conformity, whereas this decision itself may lack conformity. Besides this difficulty, few legal orders
effectively adhere to the ideal and try to implement it. In the British and American as well as in the
German tradition, the expression ‘rule of law’ or ‘Rechtsstaat’ is rather linked with a legally empty
conception of ‘justice’ and the importance of ‘judges’ in law-making (cf. Otto Pfersmann, ‘Die
normative Demokratie: Der Vorbehalt des Gesetzes und der Rechtsstaat’, in Constance Grewe,
Christoph Gusy (eds.), Französisches Staatsdenken (Baden-Baden, Nomos 2002) p. 129-145; Otto
Pfersmann and Ginevra Cerrina Feroni, ‘La “costituzione contingente”: a proposito del dibattito
sulla judicial review’, in Alessandro Torre, Luigi Volpe (eds.), La Costituzione britannica /The British
Constitution (Torino, Giappichelli 2005) p. 1271-1283)). In France, the relatively recent notion of
‘Etat de droit’ seems also mainly related to the weight of ‘judges’ in legal concretisation. The point
is that neither the concept of ‘judge’, nor its function in the legal system is clearly identified in legal
scholarship and, of course, political discourse.
http://journals.cambridge.org Downloaded: 26 Nov 2010 IP address: 128.122.149.54
233Concrete Constitutional Review in France
On the one hand, courts are considered, along the lines of Montesquieu’s all
too famous dictum of the ‘mouth of the law’, to be transparent ‘applicators’ of
the law, i.e., as weak and passive actors who can only abide by the law. On the other
hand, they are seen as rebels who always try to overstep their ambit of competence
and to intrude into the realm of politics.25
So the public perception of the
‘judge’ is not one of confidence. Even the Constitution seems to have consecrated
this mistrust in dubbing the judicial branch ‘judicial authority’ and not, as
the argument goes, ‘judicial power’.26
This, of course, is only a terminological
issue, as the French judiciary has legal powers, i.e., competences. It might nonetheless
give a hint as to why certain competences given to courts in other systems of
constitutional review are not attributed to the French ordinary courts.
The second ideological element in the French debate is the ambiguous attraction
of a model in which citizens directly apply to a ‘Supreme Court’ on rights
issues.27
The proposed reforms in 1989 and in 1993, which were not fundamentally
different from the present reform, were discussed in terms of the introduction
of an ‘exception of unconstitutionality’ and ‘direct access’ to the Council.28
25 This fear is best expressed by the concept of ‘judicial government’ popularised in France by
Edouard Lambert in his book Le gouvernement des juges et la lutte contre la législation sociale aux États-Unis
(Paris, Marcel Giard 1921; reprint Paris, Dalloz 2004). On this topic, cf. Otto Pfersmann, ‘Le concept
de “gouvernement des juges”’, in Séverine Brondel, Nobert Foulquier, Luc Heuchling (eds.),
Gouvernement des juges et démocratie. (Paris, Publications de la Sorbonne Paris 2001) p. 37-52 ; idem,
‘Remarques sur le concept de “pouvoir judiciaire”’, in Olivier Cayla, Marie-France Renoux-Zagamé
(eds.), L’office du juge: part de souveraineté ou puissance nulle? (Paris, LGDJ 2002) p. 181-193.
26
‘Title VIII The judiciary authority’ (Arts. 64 to 66-1).
27
Cf. on this point Louis Favoreu, ‘Rapport général introductif’, Revue internationale de droit comparé,
Vol. 33 (1981) p. 255-281.
28
Cf. special issue on the ‘exception d’inconstitutionnalité’, Revue française de droit constitutionnel,
1990, no. 4. The term ‘exception of unconstitutionality’ refers to a procedural structure, in which
litigating parties require that a legislative provision not be applied by the court settling the litigation
because of unconstitutionality. Neither of the French reform projects, nor the present reform are
introducing such a setting, which characterises immediate concrete review. None of the European systems
(varieties of the Austrian system), with the exception of Portugal, integrates an exception of
unconstitutionality. The use of this terminology in French scholarship shows the confusion between
the varieties of concrete review in the European model (where judges ask the constitutional
court for a – preliminary – ruling on the constitutionality of the applicable legislation) and a certain
conception of the American model, in which the jurisdiction of litigation (there is no other one)
decides on constitutionality among all other legal questions raised by the case. This confusion may
have been the reason why the French system of concrete review attributes the initiative and quality
of primary party in the constitutional procedure to the parties who raise the question of constitutionality.
I have once called this structure ‘preliminary ruling on exception of unconstitutionality’ in
order to underscore this combination of two different conceptions (Otto Pfersmann, ‘Le renvoi
préjudiciel sur exception d’inconstitutionnalité – la nouvelle procédure de contrôle concret a posteriori
selon les articles 61-1 et 62 de la Constitution’, in Les petites affiches 2008, No. 254, p. 103-110). I
have to admit that outside this particular context, the expression remains confusing. The French
construction is not an exception of unconstitutionality, but, as I try to show, an indirect complaint.
http://journals.cambridge.org Downloaded: 26 Nov 2010 IP address: 128.122.149.54
234 Otto Pfersmann EuConst 6 (2010)
This discourse was in plain contradiction to the very facts of the matter, as there
has never ever been any concrete initiative to establish a Supreme Court along
American lines or to introduce anything resembling a German Verfassungsbeschwerde,
a Spanish amparo or an Austrian Individualantrag, let alone a Hungarian actio popularis.29
This way of presenting the proposed reforms is probably due to the wish to give
individuals something of an access to the Council, while leaving it to the ordinary
courts to decide whether access should actually be granted. Even now, one can
read that individuals have access to the Council.30
Direct constitutional complaints are requests lodged by individuals with the Constitutional
court, which examines them as to the admissibility and possibly as to the
substance. These complaints can have different objects and follow different procedures.
The fact that they are lodged by individuals should not be confused with
their legal target.31
Whereas complaints like the Austrian Beschwerde or the Spanish
amparo challenge individual administrative acts (in Austria) or also (in Spain) judicial
decisions with respect to constitutionally granted rights, the German Verfassungs-
beschwerde,32
the Belgian recours direct and the Austrian Individualantrag may target
primary legislation. When primary legislation is targeted, the admissibility conditions
are usually more restrictive. Moreover, decisions quashing statutory provisions
are rare.
29 In Hungary, every person can challenge a statutory provision without any link to concrete
litigation or personal concern (actio popularis) (Art. 32a(1) Hungarian Constitution). It bears thus an
aspect of abstract review, but it gives the individual an access to the Constitutional Court, who will
obviously indulge into a severe check of the requirements for admitting such a request for a deliberation
on the merit. Cf. e.g., Wojciech Sadurski, Rights before Courts: A Study of Constitutional Courts in
Postcommunist States of Central and Eastern Europe (Berlin, Springer 2005); idem, ‘Twenty Years After
the Transition: Constitutional Review in Central and Eastern Europe’, Legal Studies Research Paper
No. 09/69 July 2009, Sidney School of Law, accessible under: ;
Laszlo Solyom, ‘The Role of Constitutional Courts in the Transition to Democracy’, International
Sociology 18 (2003), p. 133-161.
30 To give only one recent example: Gérard Longuet and Huges Portelli say in Le Monde from 9
Feb. 2010: ‘Cette procédure, qui permettra à tout justiciable de contester la constitutionnalité d’une
loi dont il estime qu’elle porte atteinte à ses droits fondamentaux…’. While not formally false, it
induces the idea, that it is indeed every litigant who could bring his or her case to the constitutional
judge.
31 Cf. on this issue the studies edited by Otto Pfersmann on ‘L’accès des personnes à la justice
constitutionnelle: droit, pratique, politique’, Cahiers du Conseil Constitutionnel 10 (2001), with an introductory
essay ‘Le recours direct entre protection juridique et constitutionnalité objective: expériences
et perspectives’, p. 110-120.
32 The German Verfassungsbeschwerde (Art. 93, par. 4b, German Basic Law) may concern any
violation of a fundamental right by ‘public authority’ (Öffentliche Gewalt). This relief is generally used
against judicial decisions, but it may also concern legislative acts if no other way is conceivable in
order to challenge a normative act impacting directly on the legal situation of a person. The case and
procedure concerning the direct challenge of a legislative provision is determined in Art. 93(3) of
the Law on the Federal Constitutional Court.
http://journals.cambridge.org Downloaded: 26 Nov 2010 IP address: 128.122.149.54
235Concrete Constitutional Review in France
Direct constitutional complaints directly targeting legislative provisions are
abstract in the sense that admissibility conditions require that statutory provisions
as such, i.e., without their application in a specific case, have an impact on the legal
situation of the complainant and violate her or his constitutional rights. Whereas
the Austrian Beschwerde presupposes the setting of an individual legal act, the Austrian
Individualantrag presupposes the absence of any prior concrete case. The French
procedure introduces by no means a direct complaint, neither against individual
acts, nor against general and abstract norms: it requires that a possibly unconstitutional
law is applicable substantively or procedurally to the case (see infra). But
neither does it leave the task of raising a constitutional issue to the ordinary courts.
This remains the exclusive competence of the parties in the case. Moreover, only
provisions which violate fundamental rights and liberties may be challenged.
The set of categories of concrete review has thus to be enlarged with the category
of indirect constitutional complaints. A procedure falls in this new category if,
first, parties can challenge statutory provisions violating their fundamental rights
in the context of a concrete case and, second, the request (in the French terminology
the ‘question’) is not lodged by one of the parties, but transmitted by the ordinary
court to the constitutional judge after a check of the fulfilment of the relevant
conditions. It combines elements of (traditional) concrete review – the courts
lodge the question (even though it is not their own) – with elements of constitutional
complaint procedures – the person to which a statutory norm applies purportedly
violating a constitutional right, challenges it, although not directly, before
the constitutional judge. Procedures are separated, but the review is subjective and
thus limited. And the question as to who is considered a party in the original litigation
will also shape the profile of subjective concrete review in the new French
setting.
Advocative subjective concrete review
One can consider still another hypothesis. The right to raise an issue of unconstitutionality
can be granted to an organ, which is neither an ordinary judge, nor one
of the parties in the original litigation directly affected by a violation of fundamental
rights. The task of this organ is to act on behalf of the directly affected
person. This competence can be exclusive or shared. In Spain, the Defensor del
pueblo and the Ministerio fiscal (Article 124 Spanish Constitution) may play this role
in the amparo referral (Article 162(2)).
In France, the advocative function exists already in abstract review, insofar as
the authorities entitled to refer acts may – and often do – raise fundamental rights
issues. The new concrete review procedure also holds an advocative element, as
the Public Ministry is entitled to raise a constitutional question if it is a party to the
original litigation. Since it, as an organ pursuing the abstract interests of the state,
http://journals.cambridge.org Downloaded: 26 Nov 2010 IP address: 128.122.149.54
236 Otto Pfersmann EuConst 6 (2010)
has obviously no subjective rights, its function as complainant in matters of fundamental
rights is advocative. This may cause a problem as it may simultaneously
act in order to restrict fundamental rights in the course of a criminal procedure.
We will come back to this.
The French reform: Preliminary ruling on individual request
after judicial examination
The French version of indirect constitutional complaint articulates three steps: a)
the question raised by a litigating party; b) the single or, as the case may be, double
transmission decision by an ordinary court; c) the ruling by the Constitutional Council
on the constitutionality issue. In doing so, the French Constitution introduces
a new second-order fundamental right and a distinctive two-step check by the
competent ordinary courts. It moreover grants large competences to the Constitutional
Council once it is in charge of the constitutional question.
A weak second-order fundamental right: limited domain and extended participation
First-order fundamental rights are structures which encompass the following four elements:
a constitutionally granted permission to act, the qualification as defective
of norms of lower rank in contradiction with such an exigency, a judicially organised
organ competent to quash such defective structures, and at least one organ competent
to raise the issue before this court.33
This concept of fundamental rights is
minimalist and concentrates on the fact that certain rights are fundamental not in
the sense that they are subjectively important, but because they are granted by the
most fundamental legal structure within a given system, i.e., the formal constitution.
This concept also implies that rights have no legal relevance if they cannot
be legal reasons for quashing legislation that violates them.
However, the concept leaves open the issue of who is concretely entitled to
raise such a violation before a (constitutional) court. It might be the person whose
right is violated, but it also might be someone else. Indeed, the competence to
raise the issue of a fundamental right’s violation has to be distinguished from the
right itself. This is a conceptual and not terminological issue: if one wants, one
can use other terms. There is, however, a need to distinguish these issues, even
though the German and the Austrian system, for instance, terminologically and
functionally equate constitutional rights with the rights that may be invoked in the
various forms of constitutional complaints. But even the German system with its
33 For a further elaboration of this comparative concept of ‘fundamental right’, cf. Otto
Pfersmann in Louis Favoreu (ed.), Droit des libertés fundamentales (Paris, Dalloz 2009, 5th edn.), chap.
‘Esquisse d’une théorie générale des droits fondamentaux’.
http://journals.cambridge.org Downloaded: 26 Nov 2010 IP address: 128.122.149.54
237Concrete Constitutional Review in France
strong conception of rights has not always had this particular property, as the
constitutional complaint was introduced into the Grundgesetz only in 1969, twenty
years after the same Grundgesetz entered into force with a list of ‘fundamental
rights’, which then lacked this specific kind of protection. Italy and France, for
different reasons, both have a set of constitutionally granted rights, but their beneficiaries
were not entitled to address the constitutional court and in Italy this
probably will not change in the near future.
A fundamental right allowing for a challenge to the violation of a first-order
fundamental right is a second-order fundamental right. For quite different reasons, France
and Italy had fundamental rights without second-order fundamental rights: the
organs entitled to refer a constitutional question to the constitutional judge were
‘political authorities’ in France and the ordinary courts in Italy. In Italy, however,
as said before, the litigants already had a weak second-order right: they could and
can raise constitutional issues, even those not related to their constitutional rights,
before an ordinary court, but not bring the case to the constitutional court them-
selves.34
The French reform establishes a slightly stronger, but still weak second-order
fundamental right. On the one hand, the parties and only the parties to concrete
litigation can raise a constitutional issue. On the other hand, the domain of review
is by necessity limited: the parties can only challenge provisions in contradiction
with first-order fundamental rights. This concerns therefore not a variety of objective
concrete reviews, but a subjective fundamental rights claim premised on its
relevance in concrete litigation. The initiative of a litigant is the necessary condition
of constitutional review a posteriori in the French system, although it is not the
sufficient condition. In this sense, it remains a – relatively – weak right. But this
weakness is also found in other fundamental rights systems, which are apparently
stronger. The right to lodge a Verfassungsbeschwerde in Germany is by no means a
right to have the question examined on the merits: the relevance of the case is
checked by a commission of three judges who may refuse to take the case by
unanimous vote without justification.35
And 98-99% of the complaints are actually
rejected every year. Under these premises, the new French procedure only gives
the exclusive right to have one’s case examined for an eventual transmission.
There is, however, an element, which exceeds this strict conception of a second-order
fundamental right. As said before, in its capacity as a party, the ‘Public
Ministry’ may also raise constitutional questions in cases in which it represents the
state either as public prosecutor in criminal cases or the interests of the state in
civil procedure (if it is not a party, it has to be informed in order to be able to
34 Constitutional law of 1 Feb. 1948, Art. 1; Art. 23, law no. 87 of 11 March 1957.
35 Art. 93d of the Law on the Federal constitutional court.
http://journals.cambridge.org Downloaded: 26 Nov 2010 IP address: 128.122.149.54
238 Otto Pfersmann EuConst 6 (2010)
present its observations).36
If the Public Ministry as a party, probably mainly in
criminal procedures, raises constitutional questions regarding rights and liberties,
these are not its own, as it as representative of the state has no fundamental rights
and liberties. In criminal procedures, the Public Ministry therefore not only has
the duty to assure the interests of the public order to have purported criminals
convicted, but also the delicate and discretionary task to raise constitutional questions
when the rights of the prosecuted persons are violated by applicable legislation.
Viewed thus, there is convergence with the more traditional structures of concrete
review, in which the courts exercise the mission of raising issues of constitutionality
in the interest of the rule of law. This shows two ways of organising
competing interests: the ordinary courts always have to remain impartial in the
original case, but in classical, i.e., objective concrete review, they may be partial
and become a party in constitutional litigation. In subjective concrete review, i.e.,
indirect complaints, courts are not parties to constitutional litigation; rather, they
solely judge the conditions of transmissibility. The extension of the right to raise a
constitutional question to the Public Ministry combines this latter structure with
advocative subjective review.
This may be considered problematic: the same organ which is in charge of
having someone deprived of the use of a fundamental right, acts in favour of the
defence of fundamental rights, as it will be the party who has to make the case for
the unconstitutionality of a legislative provision infringing the rights of the person
it accuses. These opposite functions raise a question not of impartiality, but
of coherently organised partiality.
Double selective transmission
Within all European varieties of concrete review, the French case is certainly the
most restrictive. It seems to be the consequence of the construction of the procedure
as an indirect constitutional complaint.
Depending on whether the question is raised for the first time on appeal, in
cassation or otherwise directly before the highest court in its order, there will be
one or two examinations as to whether the claim meets the following three conditions
(Article 23-2 and 23-4 of the organic law): 1) the contested provision is
applicable substantively or procedurally to the case or constitutes the foundation
of a criminal inquiry; 2) the relevant provision has not yet been declared in conformity
with the Constitution in the motives and in the operative part of a decision
of the Constitutional Council; 3) the question is not deprived of seriousness.
If the constitutional complaint is raised before a lower ordinary court, it has to
decide in the ‘shortest period of time’, that is, in fact, without any precise time
36
Cf. Art. 421 of the Code of civil procedure.
http://journals.cambridge.org Downloaded: 26 Nov 2010 IP address: 128.122.149.54
239Concrete Constitutional Review in France
limit, to transmit it to the Council of State or the Court of cassation, depending
on the nature of the case. The supreme tribunal concerned then has three months
to transmit the question to the Constitutional Council. It has to do so if the first
two above-mentioned conditions are met and if 3) the question is new or is not
deprived of seriousness. As indicated, an indirect constitutional complaint can
also be raised for the first time before the supreme tribunals. In this case, they
have three months to decide under the same conditions.
When compared with other systems of concrete review, the first condition is
more generous. A norm may be applicable to the case or to the procedure without
being, strictly speaking, conditional for the ruling on the case. This provision stems
from a parliamentary amendment – the governmental project was more in line
with the traditional conception of concrete review as ‘prejudicial’ for the ruling
on the merit and this is often considered to exclude questions concerning the
applicable procedure.
The second condition is, in my view at least, the most distinctive of the French
conception. It aims at excluding bis in idem, i.e., a challenge to an already settled
case (res judicata). For reasons to which we revert later, the determination of the
ambit of settled cases is problematic. According to the organic law, a case is ruled
when a provision has been considered in conformity with the Constitution in the
operative part (‘dispositif’) and in the justification (‘motifs’) of a previous decision
of the Council. The Constitutional Council, which had to examine the constitutionality
of this provision, ruled without any argument that this prevented a challenge
to a res judicata and was therefore in conformity with the Constitution. The
organic law attributes hence a normative character to the justification of a constitutional
decision.
There is a second interesting element in this condition. A question may be
raised even though covered by res judicata when there is a ‘change in circumstances’.
This concerns, on the face of it, a change in the legal, i.e., constitutional environment.
But according to some, it can also concern factual changes.37
What these
might be, is by no means clear and leaves ample room for discretion.
The third condition leaves extensive discretion to the transmitting court. What
is considered to be ‘serious’ is obviously open to the most opposite appreciations.
That something might be considered ‘new or serious’, as the condition reads for
the highest tribunals, is even more indeterminate, as it includes an alternative that
extends the domain of discretion. Practice, mainly by the highest tribunals, will
37
Marc Guillaume, Secretary General of the Constitutional Council, holds this view in: ‘La
question prioritaire de constitutionnalité’, to appear in: Justice et cassation, revue annuelle des avocats au
Conseil d’État et à la Cour de cassation, 2010 (accessible at )
p. 14.
http://journals.cambridge.org Downloaded: 26 Nov 2010 IP address: 128.122.149.54
240 Otto Pfersmann EuConst 6 (2010)
show whether or not this discretion will be used as an obstacle against the transmission
of otherwise relevant challenges to the Council.
A further aspect ought to be briefly mentioned. All regulations concerning
concrete review have to settle the issue as to whether the court in the original case
may proceed with the case or has to suspend it. The French setting is differentiated
and complex. Normally, the courts have to suspend the action, but the organic
law introduces three exceptions. First, if there are precise time-limits for
(urgent) decisions, they have to be met. Second, if the person concerned by the
procedure is deprived of his liberty, the issue of liberty can/has to be decided
without awaiting the decision on the issue of constitutionality. Third, ordinary
courts may decide on issues which, if left undecided, could cause irreparable damage.
Hence, it is not out of the question that the original case is definitely settled
before the constitutional question is itself definitely settled. In its decision on the
organic law, the Constitutional Council introduced a reservation holding that in
such a situation, the person concerned has the right to start a new procedure
before the ordinary court in order to have the constitutional ruling taken into
account. The Council thus added an alternative provision to the law.
Determinative decision
According to the new provisions of the Constitution, a legislative provision declared
unconstitutional is ‘abrogated’:
A provision declared unconstitutional on the basis of Article 61-1 shall be abrogated
as of the publication of the said decision of the Constitutional Council or as
of a subsequent date determined by said decision. The Constitutional Council
shall determine the conditions and the limits according to which the effects produced
by the provision shall be liable to challenge.
The organic law does not add any new and more concrete elements to this general
provision. It gives the Council ample discretion as to the time-effects of an annulment.
There is not even an upper limit to the postponement of the moment in
which an unconstitutional provision will have to loose legal effect. And it is again
the Council that rules on the concrete consequences of the elimination of an
unconstitutional provision. The finely tuned mechanisms developed in other countries,
especially Austria,38
have not in the least been taken into account. The provision
simply leaves it up to the Council to rule on the concrete consequences of
a declaration of unconstitutionality, for instance in order to give the legislature
enough time to replace the quashed provision. This further enhances its role as a
‘positive legislator’.
38 Cf. supra n. 18.
http://journals.cambridge.org Downloaded: 26 Nov 2010 IP address: 128.122.149.54
241Concrete Constitutional Review in France
The French constitutional common law
‘The force of res judicata is attached not only to the operative part (“dispositif”),
but also to the reasons which constitute its necessary foundation’ one can read in
a major decision by the Constitutional Council.39
Similar views, perhaps less explicitly
stated, are found in the decisions of many other courts and the writings of
scholars. The organic law concretising Article 61-1 of the Constitution now explicitly
states in its Article 23-2:
The tribunal states without delay with a reasoned decision on the transmission of
the priority question of constitutionality to the Council of State or the Court of
cassation. The transmission is operated if the following conditions are met: (…)
2nd
The question has not yet been declared in conformity with the Constitution in
the motives and the operative part, except a change of circumstances, by the Constitutional
Council (…). (emphasis added).
As said before, the Constitutional Council in its decision of 3 December 2009,
without any reasoning, considered that this provision meant to prevent a new
exam of a question already decided by the Council, except in the case of a ‘change
of circumstances’, and was thus ‘not contrary to the Constitution’ (15th
consider-
ation).
To this time, in France the setting on a par of motives and operative part was
just a jurisprudential fact and a scholarly opinion.40
The organic law gives it not
only a general normative foundation, but also makes a declaration of conformity
in the motives an obstacle for the transmission of a constitutional question. The
Council could have considered that this was unconstitutional. It did not. The extension
of the normative quality to legal arguments is now part of positive organic
legislation.
This raises highly perplexing issues. It confuses two things, which are distinct
by their very nature. It therefore promotes one thing (normativity) to the detriment
of the other (explanation). This technically transforms (or boosts the transformation
of) the French legal system into a specific form of common law, which
weakens the rule of law while purporting to enhance it.
Norm and reason
The identity statement is often taken as doctrinal evidence: the decision and its
main reasons are one. But are they? Most developed legal systems know some
39 French Constitutional Council Decision n° 62-18 L, 16 Jan. 1962.
40 Cf. Otto Pfersmann, ‘Relativité de l’autonomie ontologique, confirmation de l’autonomie
disciplinaire institutionnelle, paradoxe de l’hétéronomie épistémologique’, in Bertrand Mathieu (ed.),
1958-2008. Cinquantième anniversaire de la Constitution française (Paris, Dalloz 2008) p. 527-544.
http://journals.cambridge.org Downloaded: 26 Nov 2010 IP address: 128.122.149.54
242 Otto Pfersmann EuConst 6 (2010)
form of constitutional review and require decisions to be reasoned. Constitutional
review is a powerful tool for restructuring the legal system according to
normative standards set forth at the highest level of lawmaking. The reasoning of
decisions delivered in such matters has thus a particular importance as it provides,
if not a cogent justification for a particular ruling, at least the reasons purportedly
showing the path from constitutional requirements to specific outcomes. However,
the relationship between decisions and reasons for decisions is far from evident.
And the issue may even become important in the evolution of contemporary
legal systems. Merging rulings and reasons may entail a major increase in the indeterminacy
of legal orders and correspondingly reduce the normativity of the Constitution,
as I shall try to show.
Rulings of constitutional jurisdictions are usually justified, although there may
be cases where they settle a case without providing any reasons. This happens, for
instance, when the German Federal Constitutional Court refuses to hear an individual
complaint, either because the necessary admissibility conditions are not
met or because the constitutional problem is not considered to be sufficiently
relevant. In such cases, three judges of the Court may decide unanimously without
giving any reason to the claimant. But unreasoned decisions are not allowed
when primary legislation is annulled. All systems of constitutional justice require
that such decisions be ‘reasoned’ or ‘justified’ or ‘motivated’.
The obligation to justify a decision is not simply a ‘practice’ that courts may or
may not follow; it is a legal and often differentiated obligation.41
Of course, the
relevant rules may present a certain degree of indeterminacy. When a statute simply
says that decisions have to be ‘justified’, this does not say much about how a
court ought to proceed when doing so. Leaving aside the difficult question as to
whether there are normative standards concerning the level of elaboration of a
‘proof’, the form of a justification may be entirely left in the dark. Would it, for
instance, be considered legally admissible for a French court to write in a continuous
set of sentences, instead of a single sentence broken into highly complex subsentences,
linked by a present participle (‘considering that…’, ‘considering, on the
other hand, that…’, …), explicitly quoting legal scholarship or other judicial
sources? To give another example: would it be admissible for the Austrian constitutional
court to switch to a French or an American mode of ‘justification’? The
specific style may simply be a habit or practice. However, what remains is the
common and basic requirement to say why one came from certain premises to
certain consequences.
41 Examples: Austria: para. 26 Law on the Constitutional Court; France: Art. 20 of the organic
Ordinance on the Constitutional Council; Germany: para. 30, 57, 59 of the Law on the German
Federal Constitutional Court; Italy, Art. 11 and 18 of the Law on the Constitutional Court.
http://journals.cambridge.org Downloaded: 26 Nov 2010 IP address: 128.122.149.54
243Concrete Constitutional Review in France
Problems appear when one looks at the articulation of normative content and
justificatory discourse. There are strong theoretical reasons for their strict distinction
and many possibilities of arranging their unnoticed confusion. The reasons
for distinguishing decisions from their justifications are conceptual, logical and
legal.
– A judicial decision has by necessity a normative element (operative part). This is a
conceptual statement. Something which would be called ‘decision’ or ‘ ’ or
‘Urteil’ without having any normative content would not be considered a ‘decision’,
at least in this context and as a matter of convention, even though it may be
pronounced in the most prestigious manner by something called ‘Court of Justice’.
A decision settles a normative issue where generally, different competing
options are at stake. A decision may be broken down into several different steps,
some of which may not have an immediately normative content, but which settle
preliminary issues, on which the same or another court will be asked to build its
decision. Only such judicial organs and only such legal acts are relevant here.
A normative element, however weak it may be, contains at least an authorisation,
a prohibition or an obligation. This normative element exists in a certain legal
order if it has been enacted according to conditions set forth in an already valid
norm; otherwise it has no normative value in the system.42
There are two main
legal issues concerning a judicial decision: its relative validity (or relative normativity)
and its conformity.43
A legal act, which respects the necessary and sufficient conditions
for its production is relatively valid, i.e., it is valid in the system under consideration.
A valid norm then may or may not be in conformity with other legal standards,
made substantially binding on the said norm. As the legal system organises the
validity of prescriptive acts as legal, it organises the consequences of an eventual
lack of conformity.
A norm stating that some action ought to be performed (or is authorised or
prohibited) is neither true nor false. It does not say anything about how things are
in the real world and cannot be tested against any reality. Of course, a statement
that certain norms exist in a certain legal system may be true or false, but this is a
different matter. It is indeed a fundamental question of the most basic legal scholarship
whether a norm exists or not and how to establish the proof of its existence,
i.e., its (in-)validity. But a (true or false) argument concerning validity has to
be distinguished from its object, i.e., a norm, which is neither true nor false.
42 Validity is a recursive property. For a more elaborated presentation cf. Otto Pfersmann, ‘La
production des normes: production normative et hiérarchie des normes’, to appear in Michel Troper,
Dominique Chagnollaud (eds.), Traité international de droit constitutionnel (Paris, Dalloz 2010).
43 On this distinction, cf. Pfersmann, ‘La production des normes’, supra n. 42.
http://journals.cambridge.org Downloaded: 26 Nov 2010 IP address: 128.122.149.54
244 Otto Pfersmann EuConst 6 (2010)
44 Cf. Otto Pfersmann, ‘Le sophisme onomastique. A propos de l’interprétation de la constitution’,
in Ferdinand Melun Soucramamien (ed.), L’interprétation constitutionnelle (Collection Thèmes et
commentaires; Paris, Dalloz 2005) p. 33-60.
45 This is of course the position held for some time by Ronald Dworkin in several writings:
Taking rights seriously (Cambridge, Harvard University Press 1977); A Matter of Principle (Cambridge,
Harvard University Press 1985); Law’s Empire (Cambridge, Harvard University Press 1986).
– A justification is the set of arguments, which tends to show that one certain outcome
has a basis in existing law. As said before, most courts, especially in constitutional
matters, have to reason or justify their decisions. This is a normative
requirement and as such it may or may not be complied with in fact. The justification
may be deficient or prove something else than what appears in the normative
substance of the decision. It may correctly motivate a decision in a case where the
normative part itself bears no relation with the argument developed in the reasoning.
As a matter of fact, a justification may present different degrees of (in)accuracy:
concerning the legal sources, the facts of the case, and the argument of choice.
– The meaning of the relevant legal sources may or may not be correctly identified.
This is the main topic of a theory of legal interpretation.44
Whether the law
is written or unwritten, it has a meaning and the justification first has to identify
this meaning. By ‘meaning’ is of course not meant that a prescriptive proposition
means that one and only action ought to be taken.
The question as to what possible outcomes fall under one set of prescriptive
signs expressing legal norms is an open question. Certain traditional doctrines as
well as their recent restatements consider law to be result-determined and the task
of the judge as consisting in ‘constructing’ the relevant text in order to establish
the right answer.45
This position confuses different issues and concerning the
issue of meaning it is patently false.
It is false because the degree of indeterminacy and vagueness of norm formulations
in natural language is an open question, i.e., it is a matter of scientific semantic
investigation. In certain cases, the formulation may be so precise that the
meaning allows for only one correct application. But in general, norm formulations
are designed so as to allow for a more or less wide range of indeterminacy
and vagueness. In such a situation, there is still one meaning of a norm formulation,
but it consists of a whole range of elements. Interpretation as analysis of
meaning cannot do more than provide for the entire range of meanings and it can
neither exclude nor consecrate one specific element of the meaning-range under
the title of ‘interpretation’. Choosing another terminology does not change this.
It is indifferent as to whether one says that interpretation establishes a set of
meanings or that the meaning of a set of prescriptive sentences encompasses a
whole range of elements due to vagueness and indeterminacy.
http://journals.cambridge.org Downloaded: 26 Nov 2010 IP address: 128.122.149.54
245Concrete Constitutional Review in France
46 Sometimes, even true justified knowledge may of course be only partly justified, as in the
famous Gettier-paradoxes (cf. Edmund Gettier, ‘Is Justified True Belief Knowledge?’, Analysis 23
(1963), p. 121-123). The point is that only a more elaborated theory of justification may be able to
resolve the problems arising from problems with prima-facie correct justifications.
Contrary to normal adjudication, constitutional adjudication is concerned with
the question of the conformity of statutory norms, i.e., provisions of primary
legislation, with constitutional norms. This leaves in principle much less space for
the establishment of relevant facts, as the relevant facts in such matters are precisely
the norms at stake and the normative relation between them.
– Interpretation is true or false. True if the range of meaning is correctly established,
false otherwise. This raises the most difficult and disputed epistemological
issues, as the question of what makes an account of meaning correct is one of the
most disputed philosophical problems. I can however leave that question aside
here, as long as I do not have to run against the contention that establishing the
meaning is totally impossible. If it were so, there would be no point in enacting
laws or judicial rulings, nor even in using language for anything for which a relative
degree of precision is required. Even fundamentally antagonistic views on interpretation
share the idea that at least some access to the meaning of the prescriptive
sentences used to formulate constitutional or legislative (or other) norms is
possible, even though the degree of accuracy may be much lower than in other
domains of knowledge. With this proviso, interpretation is indeed true or false,
i.e., it reaches or it fails to reach the correct range of meanings.
One can at this point distinguish the truth or falseness of the premises from
the truth or falseness of the conclusions. Interpretation is, like any other form of
investigation, a set of propositions linked through logical operations. If the assumptions
are false, the consequences will be false, even though the logical operations
have been correctly applied. And it may be that the premises are false and
that yet the reasoning reaches a correct conclusion through an error in the logical
calculus. We may arrive at true conclusions, though not at true justified knowl-
edge.46
– Once meaning is established, a justification provides reasons for the choice of a
certain outcome. This is again an element of the legal obligation of the court to
provide the reasons for issuing one certain ruling with normative impact. This
latter element of the decision clearly has a different status than interpretation. It
presupposes interpretation and has to keep within the borders established by it.
Once interpretation has defined the range of semantic indeterminacy and vagueness
of legal prescriptions, or in other words, the array of discretion, the court has
to state how it uses this discretion to settle the case, i.e., it has to make a choice
http://journals.cambridge.org Downloaded: 26 Nov 2010 IP address: 128.122.149.54
246 Otto Pfersmann EuConst 6 (2010)
47 ‘Purposive interpretation’ tries to claim the contrary. This approach is not convincing. If it
appears that the relevant norms allow for deciding in another way than is prima-facie established.
The case is trivial (though it may it may obviously be highly complex in detail), as it just shows that
secunda facie the meaning is broader than it first appeared. But if the premise is that a judge under no
circumstances is allowed to make his ideas of what the law ought to be prevail on the legally determined
set of possible outcomes, the claim that it is nonetheless legally possible or even desirable
that he decides what the law in his eyes ought to be, is simply false.
48 It would be highly difficult to manage this rule. The technique of law consists normally
precisely in leaving difficult epistemological questions outside the ambit of validity, for which only
blunt formal conditions are required, whereas the more thin questions are left over for the appreciation
of conformity.
among possible outcomes and it has to provide arguments justifying this choice.
An argument aimed at justifying an outcome which falls outside the range established
by interpretation is false, as it states that something is authorised by law, of
which it was formerly shown that this outcome is excluded from the allowed out-
comes.47
All other outcomes are in this large sense technically ‘true’, even though
they are exclusive of each other. Of course, the quality of the reasoning may be
highly different from case to case and the result may be more or less convincing.
But this is legally indifferent.
An interesting property of choice-reasoning is that it allows – and thus requires
– the courts to introduce premises which are not related to legal obligations.
Insofar as they by hypothesis are not bound to choose one particular outcome
among the admissible outcomes, but are obliged indeed to choose one of them,
the reasons for the choice cannot themselves be a legal standard (otherwise there
would be no choice, at least in this respect, contrary to the hypothesis). Traditional
doctrines find this difficult to accept. They deem that the courts have a legal duty
to produce the ‘just’ result.
– There is no logical link between true interpretation and correct choice, i.e., the
justification, and the decision. It may be, as a matter of legal requirement, that a
decision is valid only if the justification is ‘correct’, that is, if the meaning is correctly
established and the steps towards the choice of the outcome are argued
without overstepping discretion or deficiencies in reasoning. But even so, there is
no logical link between the justification and the normative part of the decision.
The link is legal: the law simply conditions the validity of the norm ‘decision’ on
the presence of certain justifying elements in the act.48
This does not depend on the way in which decisions are presented. A judgment
may contain a part entitled ‘sources’, another entitled ‘facts’, and a third one entitled
‘held’, but the content of these parts may not have any relation to what the
title announces. The relevant question is therefore whether and which part of a
judgment contains a normative issue and which part, if any, the justification. If
both these elements exist, however named, then no logical link exists between
http://journals.cambridge.org Downloaded: 26 Nov 2010 IP address: 128.122.149.54
247Concrete Constitutional Review in France
them. The normative part is valid because it complies with the normative conditions
presiding its production; the explicative or justificatory part is legally correct
because it respects the substantive requirements guiding its content. And this is
the case because the interpretation of the relevant applicable norms is true and
the choice left over to the deciding body is exercised within the determined bound-
aries.
– As a legally normative statement, a decision depends in terms of validity on
other norms already valid within the system under consideration. It is an element
of a dynamic system, i.e., a system, in which nothing is valid when it is not enacted
according to an already existing norm of norm production. However, validity in
terms of reasoning is another matter. A justification is valid in terms of reasoning,
if it either truly expresses the meaning of a normative statement or develops the
consequences of it according to rules of formal reasoning, or if it introduces
propositions concerning choices within the ambit of legally admissible choices
and their logical consequences. In other words, the validity in terms of reasoning
does not depend on rules of production. Reasoning is epistemologically and/or
logically valid, whether it is explicitly formulated or not.
– A judicial decision hence bears by necessity a normative part, which is not explicative, and
possibly an explicative part, which is not normative. If this simple and elementary distinction
is correct, it follows that a judicial decision, which would not present some
sort of distinction between a normative and an explanatory part would not be
justified.
Indeed, if in a decision everything were normative, this would imply by necessity
that nothing is explained in the sense that from a set of testable true premises
certain conclusions logically – hence again in a testable manner – follow, nor in the
sense that the reasons for legally undetermined choice would be revealed and
their consequences developed. The opposite obviously holds as well, but is not
relevant here. If there were only justifications for a certain normative outcome
without a certain normative outcome, it would not be a judicial decision at all,
however interesting and convincing the reasoning may appear in other respects. It
would just be a piece of legal scholarship in the form of a judicial reasoning.
The paradoxical weakening of the rule of law
The problem, then, amounts to the following. If reasons are transformed into
norms because they cannot be both reasons and norms, then constitutional decisions
(where the same applies to other judicial decisions as well) will not be justified,
in other words they will be arbitrary.
Instead of enhancing the Rule of Law, which requires that norm production,
especially by courts, ought to be justified, it weakens justification, as reasoning are
transformed into normative statements.
http://journals.cambridge.org Downloaded: 26 Nov 2010 IP address: 128.122.149.54
248 Otto Pfersmann EuConst 6 (2010)
49 The constitutional question has ‘priority’ over the problem of conformity with ‘regularly
ratified and applied’ international treaties, but only to the extent to which both issues been raised by
the parties. Priority remains thus relative to the choice made by the parties and of course the good
will of the highest jurisdictions to transmit the question to the Constitutional Council.
Second, it weakens the Rule of Law insofar as it introduces norm statements,
which are not conceived of as norm statements. ‘Motivations’ are conceived of as
explanations and justifications; they present themselves in the garments of argument.
Instead, Article 23-2 of the organic act transforms those statements expressed
and written as arguments into the formulation of secondary constitutional
provisions, without, of course, explicitly formalising justifications as ‘secondary
constitutional acts’. This raises several difficulties. First, these provisions are difficult
to read in themselves and require a high amount of interpretive reconstruction.
Second, they introduce a high amount of indeterminacy. Third, they are by
no means systematically organised. Instead of structuring norm production and
the check of norm application, the Constitution becomes difficult to identify and
difficult to understand. And this new, barely intelligible Constitution develops in
an unorganised manner adjacent to the old Constitution, which continues, nonetheless,
to claim it is the only Constitution.
The future of French (and other) constitutional justice
French constitutional justice has proven to be very strong compared to other
European legal systems, as the Constitutional Council can rule on entire pieces of
legislation, instead of only on particular provisions in a particular case. It is exactly
the latter narrow conception, characteristic of a posteriori review, that the reform
requires the Council to learn and to adapt. The parties and their councils will have
to choose between international rights standards applicable directly by an ordinary
judge and constitutional questions with far-reaching consequences.49
The
impact of the reform will be determined by the courts, and especially by the Court
of Cassation and the Council of State: it is at their discretion whether the Constitutional
Council will have to rule on important issues. It also depends on the Constitutional
Council itself, as it decides on the (temporal) effects of annulments.
This makes for a lot of variables. Only time and empirical observation will tell
how the new procedure will work in practice.
Finally there is the structural transformation of the French constitutional system
into a kind of a common law. The consecration of case-law as secondary
constitutional law cannot but introduce more indeterminacy. This in turn will not
only increase indeterminacy in further decisions, but also the difficulties to a scholarly
presentation of the system.