Lisbon Before The Courts Comparative Perspectives

96 Mattias Wendel EuConst 7 (2011)
Lisbon Before the Courts:

Comparative Perspectives
Mattias Wendel*
Lisbon decisions of the constitutional courts in Austria, the Czech Republic, France,
Germany, Hungary, Latvia and Poland from a comparative perspective – Democ-
racy, sovereignty and identity – Permeability of national and supranational law –
Ratification of the Lisbon Treaty and its constitutional foundations – Procedural
background and legal outcome of the Lisbon decisions – Differences of institu-
tional self-conception – Parliamentary responsibility for integration – Prior parlia-
mentary assent to the future application of ‘dynamic treaty provisions’ – Different
conceptions of national and multi-levelled democracy – Popular vote – Constitu-
tional limits to European integration – Limits indicating the necessity of an amend-
ment and limits protecting the inalienable substantial core of a constitutional order
– German Bundesverfassungsgericht only court in Europe spelling out an eternity
clause in a detailed, catalogue style manner – Judicial restraint – Ultra vires and
identity review – Article 4.2 TEU as an integration clause of EU law and not a
derogation clause – Comparative dialectics
Introduction
The constitutional foundations of European integration have been subject to far-
reaching transformation. The entry into force of the Lisbon Treaty on 1 December
2009 is a milestone in this ongoing journey. However, the reform of European
Union treaty law is just one, albeit important stone in the game. Focusing exclu-
sively on the treaty changes does not capture the whole picture, for the remarkable
development of national constitutional law has shaped and conditioned the reform
of European constitutional law just as well. The activities on the Union and on
* Dr. iur. (Humboldt-University Berlin), Maîtrise en droit (Paris 1), research assistant and lec-
turer, Walter-Hallstein-Institute for European Constitutional Law (WHI), Humboldt-University
Berlin, contact: [email protected]. This article goes back to a paper presented at
the 8th conference of the European Constitutional Law Network (ECLN) 2010 in Madrid. For
helpful suggestions on earlier drafts I would like to thank Alexander Daniel, Edgar Lenski and Jan-
Herman Reestman.
European Constitutional Law Review, 7: 96–137, 2011

© 2011 T.M.C.ASSER PRESS and Contributors doi:10.1017/S1574019611100061
Lisbon Before the Courts: Comparative Perspectives 97
the member state levels were closely interrelated and may literally be told as a
story of ‘multilevel-constitutionalism in action.’1
A considerable part of today’s national constitutional law relating to the Euro-
pean Union has found its current shape in recent times. Many of the Central and
Eastern European countries that joined the Union in 2004 and 2007 enacted new
integration clauses in their constitutions providing the normative basis for Union
membership.2 However, founding members such as France and Germany as well
as ‘old’ member states such as Ireland and Portugal also passed important EU-
related amendments. Here, the provisions regulating the constitutional permeabil-
ity3 for supranational law were significantly reframed in the course of the
ratification of the Constitutional Treaty and the Lisbon Treaty.4 This process of
adjustment is continuing. In July 2010 the Austrian legislation passed a detailed
constitutional amendment with regard to parliamentary rights in EU matters.5 In
Sweden, the modernization of EU related articles is expected to come into force
on 1 January 2011, as a part of a major constitutional reform package.6 In other
countries, there are calls for the introduction of explicit EU provisions in the
constitution as well.7
Alongside the textual evolution, national supreme jurisdictions all over Europe
have delivered an unprecedented series of landmark decisions within a relatively
short period of time. These decisions address key questions of European constitu-
tionalism. Just recall the declaration of the Spanish Constitutional Tribunal on
the Constitutional Treaty in 2004 with its already famous distinction between
¹ I. Pernice, ‘The Treaty of Lisbon: Multilevel constitutionalism in Action’, 15 CJEL (2009)

p. 349 et seq.
² See on that A.E. Kellermann et al. (eds.), EU-Enlargement – The Constitutional Impact at EU
and National Level (T.M.C. Asser Press 2001); id. et al. (eds.), The Impact of EU Accession on the
Legal Orders of New Member States and (Pre-)Candidate Countries (T.M.C. Asser Press 2006);
A. Albi, EU Enlargement and the Constitutions of Central and Eastern Europe (Cambridge University
Press 2005) p. 67-121; id., ‘“Europe” Articles in the Constitutions of Central and Eastern Euro-
pean Countries’, 42 CMLRev (2005) p. 399 et seq.
³ For the concept of constitutional permeability, see M. Wendel, Permeabilität im europäischen
Verfassungsrecht (Mohr Siebeck 2011, forthcoming), ch. 1.
4 For a systematic analysis of integration clauses in the constitutions of the EU Member States
cf. Wendel, supra n. 3, chs. 4-11. For an overview, see C. Grabenwarter, ‘National Constitutional
Law Relating to the European Union’, in A. v. Bogdandy and J. Bast (eds.), Principles of European
Constitutional Law, 2nd edn. (Hart 2009) p. 83 et seq.
5 Federal constitutional law, Austrian federal law gazette I No 57/2010.
6 Bill No. 2009/10:80. The bill includes the introduction of a general clause indicating Sweden’s
Membership in the EU (future Ch. 1 § 10 of the Swedish ‘Form of Government’) as well as the
reform of the existing integration clause (currently Ch. 10 § 5, in future Ch. 10 § 6). For informa-
tion about the reform I would like to thank Carl Fredrik Bergström.
7 An example is Spain, where the Spanish State Council (Consejo de Estado) pleaded as early as
2006 for the introduction of a new and explicit ‘Europe-clause’ in its opinion of 16 Feb. 2006, No.
E 1/2005, available at <www.consejo-estado.es/pdf/modificaciones%20constitucion%20esp.pdf>.
primacy (primacía) and supremacy (supremacía).8 Similarly, the decision of the

Polish Constitutional Tribunal on the accession treaty in 2005,9 the sugar quota
cases in Hungary, Estonia and the Czech Republic from 2004 to 200610 and the
Arrest Warrant decisions in Poland, Germany, Cyprus and the Czech Republic
between 2005 and 2006 are worth noting.11 Not to forget about the French Con-
seil constitutionnel which, in 2004, delivered not only a leading case on the Con-
stitutional Treaty,12 but also started an entirely new generation of decisions
related to the transposition of directives,13 followed by the French Conseil d’Etat
in its landmark decisions Arcelor in 200714 and Perreux in 2009.15 Another major
decision was delivered recently by the German Bundesverfassungsgericht. By order
of 6 July 2010 in the case of Honeywell it has set up important procedural and
substantive limits to the exercise of ultra vires review in Germany.16
8 Spanish CT, Case 1/2004 Constitutional Treaty, declaration of 13 Dec. 2004, with case notes
of F. Castillo de la Torre, 42 CMLRev (2005) p. 1169 et seq., C.B. Schutte, 1 EuConst (2005)
p. 281 et seq. and A.C. Becker, EuR (2005) p. 353 et seq.
9 Polish CT, Case K 18/04 Accession Treaty, judgment of 11 May 2005. An English summary is
available at: <www.trybunal.gov.pl/eng/summaries/documents/K_18_04_GB.pdf>. For com-
ments, see M. Bainczyk and U. Ernst, EuR (2006) p. 247 et seq.; A. Łazowski, 3 EuConst (2007)
p. 148 et seq.; S. Biernat, ‘Offene Staatlichkeit’, in A. v. Bogdandy and P.M. Huber (eds.), Ius Pu-
blicum Europaeum, Vol. 2 (C.F. Müller 2008) § 21 Polen, para. 45.
¹0 For a comparative analysis, see A. Albi, ‘Ironies in Human Rights Protection in the EU: Pre-
Accession Conditionality and Post-Accession Conundrums’, 15 ELJ (2009) p. 46, 52 et seq.; id.,
‘Supremacy of EC Law in the New Member States’, 3 EuConst (2007) p. 25, 48 et seq.; W. Sadur-
ski, ‘“Solange, chapter 3”: Constitutional Courts in Central Europe – Democracy – European
Union’, 14 ELJ (2008) p. 1, 6 et seq.
¹¹ See on that J. Komárek, ‘European constitutionalism and the European Arrest Warrant – in
Search of the Limits of “Contrapunctual Principles”’, 44 CMLRev (2007) p. 9, 16 et seq.; Z. Kühn,
‘The European Arrest Warrant, Third Pillar Law and National Constitutional Resistance/Accept-
ance’, 3 CYELP (2007) p. 99 et seq.
¹² French CC, Case 2004-505 DC Constitutional Treaty, decision of 19 Nov. 2004. Cf. the com-
ments of G. Carcassonne, 1 EuConst (2005) p. 293 et seq.; F. Chaltiel, 484 RMC (2005) p. 5 et seq.;
X. Magnon, 62 RFDC (2005) p. 329 et seq.; J. Roux, RDP (2005) p. 59 et seq.
¹³ French CC, Case 2004-496 DC E-Commerce, decision of 10 June 2004. See the case notes of
F.C. Mayer, EuR (2004) p. 925 et seq. (also relating to the decision on the Constitutional Treaty);
J. Dutheil de la Rochère, 42 CMLRev (2005) p. 859 et seq.; J.-H. Reestman, 1 EuConst (2005)
p. 302 et seq. French CC, Case 2006-540 DC Information Society, decision of 27 July 2006. For
comments see F. Chaltiel, RFDC (2006) p. 837 et seq. and C. Charpy, 3 EuConst (2007) p. 436,
445 et seq.
¹4 French CE, Case 287110 Ass. Arcelor, decision of 8 Feb. 2007, para. 11. Cf. the case notes of
P. Cassia, RTDE (2007) p. 406 et seq.; F. Chaltiel, RMC (2007) p. 335 et seq.; X. Magnon, RFDA
(2007) p. 578; A. Levade, RFDA (2007) p. 564, 577; C. Charpy, 3 EuConst (2007) p. 436, 440 et
seq. and p. 452 et seq.; F.C. Mayer and E. Lenski and M. Wendel, EuR (2008) p. 63 et seq.
¹5 French CE, Case 298348 Mme P, decision of 30 Oct. 2009, para. 9; cf. C. Charpy, 6 EuConst
(2010) p. 123 et seq. and C.D. Classen, EuR (2010) p. 557 et seq.
¹6 German BVerfG, Case 2 BvR 2661/06 Honeywell, order of 6 July 2010, paras. 58 et seq. The
decision was published not until 26 Aug. 2010. An English translation is available at <www.bun-
desverfassungsgericht.de/en/decisions/rs20100706_2bvr266106en.html>.
When ‘Lisbon’ was brought before the courts, it was thus in the context of a
highly dynamic evolution of European constitutional law, both at textual and
jurisprudential levels. Supreme jurisdictions of several member states – old and
new – took Lisbon as an opportunity to add major voices to this jurisprudential
choir. It is not exaggerated to claim that the three-year Lisbon saga has become
one of the most important cross-border lines of jurisprudence in the history of
European constitutionalism, not only in numbers but particularly in terms of
substance.
The first decision was issued on 20 December 2007 by the French Conseil
constitutionnel,17 followed by the order of the Austrian Constitutional Court on
30 September 2008,18 the first judgment of the Czech Constitutional Court on
26 November 2008,19 the judgment of the Latvian Constitutional Court on
7 April 2009,20 the judgment of the German Bundesverfassungsgericht on 30 June
2009,21 the second judgment of the Czech Constitutional Court on 3 November
2009,22 the judgment of the Hungarian Constitutional Court on 12 June 201023
and the second order of the Austrian Constitutional Court the very same day.24
The most recent Lisbon decision so far was delivered by the Polish Constitutional
Tribunal on 24 November 2010.25 Another case is still pending before the Danish
¹7 French CC, Case 2007-560 DC Treaty of Lisbon, decision of 20 Dec. 2007.

¹8 Austrian CC, Case SV 2/08-3 et al. Treaty of Lisbon I, order of 30 Sept. 2008.
¹9 Czech CC, Case Pl ÚS 19/08 Treaty of Lisbon I, judgment of 26 Nov. 2008. An English
translation is available at <http://angl.concourt.cz/angl_verze/doc/pl-19-08.php>. See the case note
of P. Bříza, 5 EuConst (2009) p. 143 et seq.
²0 Latvian CC, Case 2008-35-01 Treaty of Lisbon, judgment of 7 April 2009. An English trans-
lation is available at <www.satv.tiesa.gov.lv/upload/judg_2008_35.htm>.
²¹ German BVerfG, Case 2 BvE 2/08 et al. Treaty of Lisbon, judgment of 30 June 2009, BVer-
fGE 123, 267 et seq. An English translation by the BVerfG (final version) is available at: <www.
bverfg.de/entscheidungen/es20090630_2bve000208en.html>. For the multitude of comments
compare the 9 pages (sic) of bibliography in the first special issue of EuR (2010) p. 325-333. For
mainly critical assessments, see in particular the comments of D. Thym, 46 CMLRev (2009) p. 1795
et seq.; R. Bieber, 5 EuConst (2009) p. 391 et seq.; C. Schönberger, 10 GLJ (2009) p. 1201 et seq.;
D. Halberstam and C. Möllers, 10 GLJ (2009) p. 1241 et seq.; C.D. Classen, 64 JZ (2009) p. 881
et seq.; M. Jestaedt, 48 Der Staat (2009) p. 496 et seq.; U. Everling, EuR (2010) p. 91 et seq.; J.
Schwarze, EuR (2010) p. 108 et seq.; C. Tomuschat, 70 ZaöRV (2010) p. 251 et seq.; T. Eijsbouts,
6 EuConst (2010) p. 199 et seq. For more affirmative appraisals cf. F. Schorkopf, 10 GLJ (2009)
p. 1219 et seq.; D. Grimm, 5 EuConst (2009) p. 353 et seq.; K.F. Gärditz and C. Hillgruber, 64 JZ
(2009) p. 872 et seq.
²² Czech CC, Case Pl ÚS 29/09 Treaty of Lisbon II, judgment of 3 Nov. 2009. An English trans-
lation of the most important sections by J. Komárek is contained in 6 EuConst (2009) p. 345 et seq.
For the perspective of a German legal scholar, see I. Ley, 65 JZ (2010) p. 165 et seq.
²³ Hungarian CC, Case 143/2010 (VII. 14.) Treaty of Lisbon, judgment of 12 July 2010. An
English translation has not been rendered yet, except for a short press review. For translation and
important information I would like to thank warmly Adél Holdampf and Attila Vincze.
²4 Austrian CC, Case SV 1/10-9 Treaty of Lisbon II, order of 12 June 2010.
²5 Polish CT, Case K 32/09 Treaty of Lisbon, judgment of 24 Nov. 2010.
Supreme Court.26 Alongside the decisions of national supreme jurisdictions, there

were a number of important advisory opinions and reports, such as the opinion
of the Dutch Council of State of 12 September 2007 on the pre-Lisbon IGC
mandate,27 the opinion of the Danish Ministry of Justice of 4 December 200728
and the report of the British House of Lords of 13 March 2008.29
Although all of these decisions and opinions paved (or confirmed) the way for
ratification in one way or another, they reveal significant differences in proce-
dural as well as in substantial terms. This article will assess the Lisbon jurisprudence
from a comparative perspective.30 After a brief preliminary reflection about how
to compare, the analysis addresses the procedural background and gives a general
overview of the decisions. It then tackles the substantial key issues for the future
development of EU law which were raised by the decisions, particularly the de-
mands of the supreme jurisdictions regarding democratic legitimacy of EU author-
ity and the courts’ claims of constitutional limits and judicial reservations. A
concluding remark aims at a question of judicial methodology as the Lisbon deci-
sions reveal a remarkable quality of comparative dialectics between the supreme
jurisdictions of the member states.
²6 See on that J.H. Danielsen, ‘One of Many National Constraints on European Integration:
Section 20 of the Danish Constitution’, 16 EPL (2010) p. 181, 190 et seq.
²7 Dutch State Council, Case W02.07.0254/II/E Lisbon-Mandate, opinion of 12 Sept. 2007.
An English translation is available at <www.raadvanstate.nl/adviezen>. For a comment, see J. Ziller,
‘The Law and Politics of the Ratification of the Lisbon Treaty’, in S. Griller and J. Ziller (eds.), The
Lisbon Treaty (Springer 2008) p. 309, 319 et seq.
²8 Contrary to what it had said in respect to the Constitutional Treaty, the Danish Ministry of
Justice qualified the Lisbon Treaty as a treaty which did not transfer competences in the sense of the
constitutional integration clause in the Danish constitution (Art. 20) and thus could be ratified like
an ordinary treaty of public international law under Article 19 of the Danish constitution.
²9 ‘The Treaty of Lisbon: an impact assessment’, report of 13 March 2008, available at: <www.
publications.parliament.uk/pa/ld200708/ldselect/ldeucom/62/62.pdf>.
³0 Apart from the almost uncountable number of comments on the decision of the German
BVerfG, there are apparently only three contributions dealing with some of the Lisbon decisions in
a comparative perspective, cf. J.-H. Reestman, ‘The Franco-German Constitutional Divide’, 5 Eu-
Const (2009) p. 374 et seq. specifically concerning the aspect of constitutional identity; A. Weber,
‘Die Europäische Union unter Richtervorbehalt’, 65 JZ (2010) p. 157 et seq. with a comment on
the German Lisbon decision in a broader comparative perspective, and R.U. Krämer, ‘Looking
through Different Glasses at the Lisbon Treaty: The German Constitutional Court and the Czech
Constitutional Court’, in A. Fischer-Lescano et al. (eds.) ‘The German Constitutional Court’s Lis-
bon Ruling: Legal and Political Science Perspectives’, ZERP-Diskussionspapier 1/2010 (Universität
Bremen 2010) P. 11 et seq. comparing the first decision of the Czech CC with the Lisbon judgment
of the German BVerfG.
A preliminary word about comparison

Comparing must not be cherry-picking. This is why the attempt of a comparative
analysis is a challenge in the present context. Some of the judgments under review
have almost epic sizes. For instance, the judgment of the German Bundesverfas-
sungsgericht consists of 421 paragraphs, which add up to more than 140 pages in
the original print version. These circumstances make it inevitable to focus on the
essential lines of argument in order to compare. But how to separate the essential
from the nonessential? The particular difficulty here lies not so much in length,
but first and foremost in diversity and (deliberate) ambiguity.
Taking the Lisbon judgment of the Bundesverfassunggericht again as example, it
is clear that this decision can hardly be described as a monolithic product of judi-
cial reasoning. It is, rather, an expression and reconciliation of a variety of dissonant
voices within the German court’s Second Senate.31 The above-mentioned Honey-
well decision is a prime example in this respect. While the majority of the Second
Senate establishes remarkable limits to the exercise of ultra vires review,32 dissent-
ing judge Herbert Landau holds that the majority hereby ‘departs from the con-
sensus on which the Lisbon judgment was based.’33 It is interesting to see that the
merits of the case as well as the dissenting opinion refer basically to the same sec-
tions of the Lisbon judgment dealing with the principle of Europarechtsfreundlichkeit
(literally ‘friendliness towards European law’).34 These sections apparently leave
enough margin of interpretation for the judges to draw almost antithetic conclu-
sions. If the judges involved cannot agree on the precise content of the consensus
on which the Lisbon judgment was based, how should we?
Abstract notions such as the principle of friendliness towards European Law35
or the principle of sovereign statehood36 become even more problematic when it
³¹ An author has aptly compared the resulting multitude of interpretations with the Japanese
film ‘Rashomon’. All characters have experienced or suffered the same incident, but recount it
completely different. See F.C. Mayer, ‘Rashomon in Karlsruhe – A Reflection on Democracy and
Identity in the European Union’, Jean Monnet Working Paper 5/10.
³² German BVerfG Honeywell, supra n. 16 at para. 58 et seq.
³³ Ibid., para. 102.
³4 The term is translated by the BVerfG as ‘openness towards European law’. It must be doubted
that this is an appropriate translation as the German term then would have been – semantically
even more precise – framed ‘Europarechtsoffenheit’.
³5 German BVerfG Treaty of Lisbon, supra n. 21 at para. 225, 240 et seq., and 340. For possible
meanings of the concept, see A. Voßkuhle, ‘Der europäische Verfassungsgerichtsverbund’, NVwZ
(2010) p. 1, 5; id., ‘Multilevel Cooperation of the European Constitutional Courts’, 6 EuConst
(2010) p. 175 et seq.; F.C. Mayer, ‘Europarechtsfreundlichkeit und Europarechtsskepsis in der
Rechtsprechung des Bundesverfassungsgerichts’, in T. Giegerich (ed.), Der offene Verfassungsstaat des
Grundgesetzes nach 60 Jahren (Duncker & Humblot 2010) p. 237, 256 et seq.
³6 German BVerfG Treaty of Lisbon, supra n. 21 at para. 224, 228 et seq., 247 et seq., 263, 280
et seq., 299 et seq.
comes to cross-border comparison. Not only the ambiguities inherent in the con-
cepts themselves, but also the different legal cultures and traditions may cause
confusion.37 Comparison then runs the risk of getting lost in presuppositions and
preconceptions. Picking out an abstract figure of argument contained in judgment
A and seeking for possible equivalents in judgment B might end up a comparison
of cherries with bananas.
In order to achieve more conclusive results, it is thus reasonable to restrict
comparative efforts specifically to those statements which have materialised at least
to some degree in the legal outcome. As a consequence, scholarly obiter dicta with
generic statements about the nature and finality of the European Union are not
central to this comparative review, even though they were written with the inten-
tion to have a lasting impact on (academic) debate.38 Instead, the following
analysis focuses particularly on those sections of the merits that determine the
judgments’ results.
Even compared in this way, the Lisbon decisions reveal considerable discrepan-
cies. In particular, the varying premises of democratic legitimacy in multi-level-
systems entail varying conclusions regarding the constitutional requirements for
the participation of national parliaments. Further, different institutional self-
conceptions and normative ideas of sovereignty and (national) identity entail
different conceptions of constitutional limits and judicial reservations. But before
addressing these substantive questions, let us recall the procedural background of
the Lisbon cases and give a general overview over the decisions. Notably, enlighten-
ing differences become apparent even following such an examination.
Procedural background and general overview

Three procedural settings must be distinguished. First, objective treaty reviews
before ratification, secondly objective treaty reviews after ratification, and finally
individual complaints or petitions being admissible only if the complainant dem-
onstrates an individual encroachment.
Preventive treaty reviews: France and the Czech Republic

Preventive, i.e., ex ante treaty reviews were carried out in France and the Czech
Republic.
³7 For the obscurity and polysemy of the term sovereignty in a comparative context, see Wendel,
supra n. 4, ch. 3.
³8 In that sense D. Thym, ‘In the Name of Sovereign Statehood’, 46 CMLRev (2009) p. 1795,
1821.
France – necessity of a prior constitutional amendment

The only judicial body which has ruled in substance39 both on the Constitu-
tional Treaty as well as on the Lisbon Treaty is the French Conseil Constitutionnel.40
In both cases it was asked by the French president to review the compatibility of
the respective treaty with the Constitution by means of an objective, ex ante review
under Article 54 of the French Constitution.
The two decisions of the Conseil constitutionnel illustrate that the Constitu-
tional Treaty and the Lisbon Treaty are largely congruent. Not in terms of termi-
nology and symbols, as the Dutch Council of State aptly pointed out in its
advisory opinion on the IGC mandate 2007,41 but congruent certainly with regard
to their substantial implications on the member state level. In the Lisbon decision
the Conseil could thus widely refer42 to its previous decision on the Constitu-
tional Treaty in which it had decided that the authorisation to ratify the treaty
required a prior revision of the French Constitution. Not surprisingly, neither the
binding character of the Charter of Fundamental Rights nor the principle of
primacy as enshrined in Article I-6 of the Constitutional Treaty required a consti-
tutional revision.43 In this respect the decision of the Conseil was in line with the
declaration of the Spanish Constitutional Tribunal,44 the advisory opinion of the
Belgian Conseil d’Etat45 and the report of the Swedish legislative council (lagrådet),
which all did not deem a constitutional revision necessary in view of the Consti-
tutional Treaty.46
However, the Conseil constitutionnel demanded a prior constitutional amend-
ment for several other reasons: firstly the conferral of certain new competences to
the EU, secondly the introduction of supranational modes of decision-making for
competences already conferred to the Union, thirdly the introduction of the gen-
eral bridge clause (now Article 48.7 TEU) and its equivalents in specific fields and,
³9 The Austrian CC has also delivered decisions on both occasions, but rejected all remedies as
inadmissible, vide infra.
40 See the contribution of J. Dutheil de la Rochère, in I. Pernice and J.M. Beneyto Pérez (eds.),
Europe´s Constitutional Challenges in the Light of the Recent Case Law: Lisbon and Beyond (Nomos
2011 forthcoming).
4¹ Dutch State Council Lisbon-Mandate, supra n. 27, point 3.4. See on that J. Ziller, ‘The Law
and Politics of the Ratification of the Lisbon Treaty’, in S. Griller and J. Ziller (eds.), The Lisbon
Treaty (Springer 2008) p. 309, 322 et seq.
4² French CC Treaty of Lisbon, supra n. 17, in particular paras. 12, 21, 24, 26 et seq. and 29.
4³ French CC Constitutional Treaty, supra n. 12 at paras. 13 and 22.
44 Spanish CT Constitutional Treaty, supra n. 8, in particular points II-4 and II-6.
45 The advisory opinions are not published. For an analysis cf. F. Delpérée, ‘Le Conseil d’Etat de
Belgique et le traité établissant une Constitution pour l’Europe’, 21 RFDA (2005) p. 242 et seq.
46 Cf. J. Nergelius, ‘Sweden’s Possible Ratification of the EU Constitution: A Case-Study of
“Wait and See”’, in A. Albi and J. Ziller (eds.), The European Constitution and National Constitutions
( Kluwer 2007) p. 183, 187.
last but not least, the new powers given to national parliaments under EU law.47
Hence, in the aftermath of both decisions a constitutional amendment had to be
passed.48 In both cases, these amendments expressly authorised France to partici-
pate in the European Union under the conditions laid down in the Constitu-
tional Treaty and the Lisbon Treaty respectively.49 Both revisions also aimed at the
constitutional authorisation and implementation of the new rights of the French
parliament under EU law.50 Most of the new provisions were enacted under the
condition of the coming into force of the respective treaty. As the ratification of
the Constitutional Treaty failed in 2005, the major part of the provisions regard-
ing that Treaty never came into effect, while their successors regarding the Lisbon
Treaty only entered into force on 1 December 2009.51 The French revision pro-
cedure is thus a classic example of the mutual interaction and interdependence of
national and supranational constitutional law.
47 French CC Constitutional Treaty, supra n. 12 at paras. 27 et seq. and later French CC Treaty
of Lisbon, supra n. 17, in particular paras. 18 et seq. With regard to parliamentary rights, the CC
demanded in its Lisbon decision – in addition to what it had already decided in its previous decision
concerning the Constitutional Treaty – a constitutional revision also with regard to the parliamen-
tary veto right under Art. 81.3 TFEU and the subsidiarity control mechanism under the reframed
Art. 7.3 of Protocol No. 2 on the exercise of the principles of subsidiarity and proportionality. See
paras. 30-32 of the Lisbon decision.
48 Constitutional Laws No. 2005-204 of 1 March 2005 (with view to the Constitutional Treaty)
and No. 2008-103 of 4 Feb. 2008 (with view to the Lisbon Treaty). In addition, the EU related
provisions of the French constitution were amended significantly by the constitutional reform-
package contained in Constitutional Law No. 2008-724 of 23 July 2008 which aimed at the ‘mod-
ernisation of the institutions’. With Art. 61-1 this law also introduced the constitutional basis for
the new a posterior review-powers of the French CC (so-called question prioritaire de constitution-
nalité).
49 Art. 88-1. Before the coming into force of the Lisbon Treaty Art. 88-1 contained a separate
para. 2 according to which France ‘can participate’ (‘peut participer’) in the EU under the condi-
tions laid down in the Constitutional Treaty or the Lisbon treaty (see the respective first article of
the Constitutional Laws No. 2005-204 and No. 2008-103).
50 Cf. Arts. 88-4, 88-6 and 88-7 of the French Constitution. Similar to Art. 23.1a of the Ger-
man Basic Law, Art. 88-6.3 frames the subsidiarity action as a parliamentary minority right.
5¹ With the coming into force of the Lisbon Treaty, the general integration clause in Art. 88-1
was also recast, according to Art. 2 of the constitutional law 2008-103. Art. 88-1 now states that
the French Republic ‘shall participate in the European Union constituted by States which have
freely chosen to exercise some of their powers in common by virtue of the Treaty on European
Union and of the Treaty on the Functioning of the European Union, as they result from the treaty
signed in Lisbon on 13 Dec. 2007.’
Czech Republic – two decisions, no amendments

Like in France, in the Czech Republic the Lisbon Treaty was submitted to ex ante
review.52 The Czech Senate filed a petition under Article 87.2 of the Czech con-
stitution, introduced in the course of the pre-accession amendment in 2001 in
order provide a basis for the preventive review of international treaties.53 In its
voluminous landmark decision of 26 November 2008 – consisting of 218 para-
graphs – the Czech Constitutional Court found the Lisbon treaty to be compat-
ible with the Czech constitutional order.54 However, the court limited its scrutiny
to those provisions of the Lisbon Treaty expressly contested by the petitioner.55
In doing so the Czech Court left the door open for another petition and thus
allowed the political opponents of the Lisbon Treaty to initiate a second proceed-
ing. Almost a year after the first decision, on 29 September 2009, a group of
Senators filed a second petition which was apparently influenced by the reasoning
of the German Lisbon decision. In its second Lisbon judgment of 3 November
2009 the Czech Court decided that the Lisbon Treaty as a whole did not conflict
with the Czech constitutional order.56 This second judgment is particularly inter-
esting from a comparative perspective as the Czech Court expressly underlined
the principle of judicial self restraint and set a clear and articulate contrast to
central parts of its German counterpart.
Furthermore, the Czech Court tried to establish limits to a potential proce-
dural abuse of the ex ante treaty review. Referring to obligations under interna-
tional as well as domestic (constitutional) law, the Court established the requirement
to remove doubts on the constitutionality of an international treaty ‘without
undue delay.’57 Although it made perfectly clear that the second petition of the
Senators, filed about a year later than the first, did not meet this requirement, the
Court nevertheless declared the petition admissible in order to avoid ‘retroac-
tively burden[ing] the petitioners’ with this new interpretation of the relevant
procedural rules.58 Furthermore, the Court held that the Czech president had an
obligation to ratify without undue delay an international treaty duly negotiated
by the executive and approved by the democratically elected legislator. It stated
5² For the decisions of the Czech CC, see the contribution of J. Zemánek, in I. Pernice and
J.M. Beneyto Pérez (eds.), Europe´s Constitutional Challenges in the Light of the Recent Case Law:
Lisbon and Beyond (Nomos 2011 forthcoming).
5³ Constitutional Law No. 395/2001 of 18 Oct. 2001.
54 Czech CC Treaty of Lisbon I, supra n. 19.
55 Albeit including some provisions which were already in force within the framework of the
EU- and the EC-treaty, cf. ibid., paras. 75, 77 et seq. and 85-87. See on that P. Bříza, supra n. 19 at
p. 145 et seq.
56 Czech CC Treaty of Lisbon II, supra n. 22.
57 Ibid., paras. 115 et seq.
58 Ibid., para. 121.
that this obligation exists a fortiori when a treaty has been approved with the
qualified constitutional majority under the terms of the Czech integration clause
(Article 10a).59 In other words, when doubts about the constitutionality of an EU
treaty arise, only a preventive treaty review initiated within an appropriate period
of time can postpone ratification until either a decision of conformity is issued or,
in case of conflict, a constitutional amendment is passed.60
Ex post treaty reviews: Hungary and Poland

In Hungary and Poland the Lisbon treaty was reviewed within the framework of
an objective ex post review.
Hungary – ‘the use of the ex ante review would be desirable’

In Hungary ratification had been conducted without a prior review of constitu-
tionality. Although this procedure is explicitly provided for, 61 none of the poten-
tial petitioners – neither the Hungarian parliament, nor the president or the
government – filed a request. Instead, the Hungarian Constitutional Court de-
cided on the constitutionality of the Act of promulgation of the Lisbon Treaty62
within the framework of an ex post review introduced by a person acting in private
capacity (actio popularis).63 The petition emphasized that the Treaty jeopardized
the existence of Hungary as an independent and sovereign state, governed by the
rule of law.
While the Hungarian Court declared the petition admissible, it dismissed the
case on the merits.64 In dealing with the procedural point that the Hungarian
ratification had already been completed, the Court underlined that even if the
Treaty in question was declared unconstitutional, the adherence to Hungary’s legal
commitments deriving from EU membership would not be threatened. In that
hypothetical situation it would be up to the legislator to find a solution in which
the obligations arising under EU law were observed without the violation of the

6¹ Under Art. 36.1 of the Act on the Constitutional Court the Hungarian parliament, the pres-
ident and the government may request the examination of the constitutionality of provisions of the
international treaty before its confirmation.
6² Act CLXVIII of 2007.
6³ According to Art. 1 lit b) of the Act on the Constitutional Court the competence of the Con-
stitutional Court includes i.a. the ex post examination for unconstitutionality of laws.
64 Hungarian CC Treaty of Lisbon, supra n. 23. Two separate opinions (supporting the over-all
result but differing as to the grounds) and a dissenting opinion were delivered with the judgment.
While the first separate and the dissenting opinion tackled the question of admissibility in a ‘Euro-
friendly’ way, the second separate opinion highlighted the constitutional limits to the conferral of
competences and the principle of primacy.
Hungarian constitution.65 But it did not come to this worst-case scenario as the
Court found the Lisbon treaty to be entirely compatible with the constitution.
However, it gave the competent political actors a broad hint, that in case of a
major reform package like the Lisbon Treaty, the use of the ex ante treaty review
would be desirable.66
Poland – protest during the oral hearing

In relation to the use of the preventive treaty review, the situation was similar in
Poland. The President of the Polish Republic ratified the Lisbon Treaty on 9 Oc-
tober 2009 without having exercised his power to initiate an ex ante review under
Article 133.2 of the Polish Constitution.67 The parliamentary statute approving
the act of ratification had been adopted pursuant to the procedure of Article 90.2
of the Polish Constitution, which requires an even more demanding majority for
the transfer of competences than for a constitutional amendment.68 After the
ratification had been completed, a group of deputies and a group of senators filed
petitions for an ex post treaty review under Article 188 no 1 of the Polish Consti-
tution.69
In its judgment of 24 November 2010 the Polish Constitutional Tribunal found
the Lisbon Treaty to be compatible with the Polish Constitution. Taking into
account that the Polish ratification had been authorised by a qualified legislative
procedure and carried out by the president who himself had an obligation to
ensure that the Polish Constitution was respected, the Tribunal took the view that
the Lisbon Treaty enjoyed a presumption of constitutional conformity which could
not be overturned in the present case. Like its Hungarian counterpart, the Con-
stitutional Tribunal had to cope with the procedural particularities of an ex post
review and found an answer on its own with this line of argument.
In addition, the Polish Lisbon case was characterised by a procedural curiosity.
Ultimately, the Constitutional Tribunal decided only in relation to the senators’
petition, which was essentially about the constitutionality of the simplified revision
65 Ibid., point IV.2.

66 Ibid., point IV.2.2.
67 According to this provision the president, before ratifying an international agreement may
refer it to the CT with a request to adjudicate upon its conformity to the Polish constitution.
68 Both Arts. 90.2 (parliamentary procedure in case of a conferral of competences to interna-
tional organisations) and 235.4 (amendment procedure) require a two-thirds majority vote in the
chamber of deputies in the presence of at least half of the statutory number of Deputies. The differ-
ence is that Art. 90.2 also requires a two-thirds majority vote in the Senate in the presence of at least
half of the statutory number of Senators, while Art. 235.4 only requires an absolute majority of
votes in the presence of at least half of the statutory number of senators.
69 This provision establishes the competence of the Polish CT to decide on the conformity of
statutes and international agreements with the Polish Constitution.
procedure, the general regime of Union competences and the flexibility clause
under Article 352 TFEU. In contrast, the Tribunal had to drop the case relating
to the much more comprehensive petition of the group of deputies. The simple
reason for this was that their representative member had left the courtroom in
protest during the oral hearing and was therefore regarded as being absent.70
However, it is not out of the question that the same group of deputies will make
a second effort to bring the case before the Tribunal, including the review of con-
stitutionality of the new accompanying legislation which regulates the cooperation
of the Polish government and the parliament in EU affairs.
Individual complaints and petitions: Germany, Austria and Latvia

While in France, the Czech Republic, Hungary and Poland the constitutionality
of the Lisbon treaty was examined within the procedural framework of an objec-
tive treaty review, in Austria, Germany and Latvia the proceedings were initiated
by individual complaints or petitions which demand the complainant establish
(prima facie) a personal interest for review deriving from an individual encroach-
ment, such as the alleged infringement of a fundamental right.
Germany – the right to vote as a catalyst to a full-scale review

The Lisbon judgment in Germany was predominantly71 based on individual con-
stitutional complaints under Article 93.1 no. 4a of the German Basic Law. In 2005
the German Bundesverfassungsgericht had already been confronted with two com-
plaints directed against the act approving the ratification of the Constitutional
Treaty.72 But after the negative outcome of the referenda in France and the Neth-
erlands and the following period of reflection, the German Court – informally –
decided not to decide for the time being. The reason given was that it did not want
to ‘actively contribute’ to the discussion about the future of European constitu-
tional development.73 It could be argued, however, that the decision not to decide
was a political contribution as well, given the context that, at that time, it was
more than unclear whether the constitutional reform process would continue.
70 The representative of the group of deputies, deputy Antoni Macierewicz, stormed out of the
auditorium after the CT had rejected his motion to postpone the judgment until the bill regulating
the cooperation of the government and the parliament in EU affairs came into force. The CT
rejected the motion because the initial proceeding covered only the constitutionality of the Lisbon
Treaty, but not of the (future) accompanying legislation.
7¹ The intra-institutional proceedings initiated by the parliamentary group of the Left Party
against the chamber of deputies (Bundestag) were declared inadmissible to a large extent.
7² Cases 2 BvR 839/05 and 2 BvE 2/05.
7³ This argument was given in a letter by the then judge rapporteur Broß addressed to the par-
ties.
However, that may be, unlike the French Conseil constitutionnel, the Bundes-
verfassungsgericht thus only delivered a decision on the Lisbon Treaty. Like in the
previous Maastricht judgment74 the catalyst for admissibility was the right to vote
under Article 38 § 1 of the German Basic Law. The German court construes this
right dogmatically as ‘equal to a fundamental right’ and substantially as
a right to democratic self-determination, to free and equal participation in the state

authority exercised in Germany and to compliance with the principle of democracy
including the respect of the constituent power of the people. In the present combi-
nation of procedural circumstances, the review of a violation of the right to vote also
comprises encroachments on the principles which are codified in Article 79 § 3 of
the Basic Law as the identity of the constitution.75
According to the court, the complainants in the Lisbon case could thus rely on
Article 38 § 1 in order to claim a violation of the principle of democracy, the loss
of German statehood and ‘a violation of the principle of the social state.’76 The
wide interpretation of Article 38 § 1 had already been vividly criticised in the
aftermath of the Maastricht judgment77 as it enables virtually every German hav-
ing the right to vote to initiate a de facto objective review of constitutionality, al-
though this specific procedure is only open to an enumerated circle of petitioners
under Article 93 § 1 no. 2 of the Basic Law.78
What is new in the Lisbon decision is that the Bundesverfassungsgericht connects
the right to vote with the constitutional identity as a whole and, furthermore, with
the ‘respect of the constituent power of the people.’ It thus extends its scrutiny in
two ways.79 First, by invoking Article 38 § 1, a German individual can now claim
the violation of theoretically all principles protected by the eternity clause of Ar-
ticle 79 § 1, as far as he or she plausibly demonstrates a ‘necessary connection’ of
these principles with the principle of democracy.80 Hence, the German court could
not only declare admissible the allegation that the principle of democracy was
74 German BVerfG, Case 2 BvR 2134 et al. Treaty of Maastricht, judgment of 12 Oct. 1993,
BVerfGE p. 89, 155, 171 et seq.
75 German FCC Treaty of Lisbon, supra n. 21 at para. 208.
76 Ibid., para. 168. The ‘principle of the social state’ is part of the basic principles under
Art. 20.1 and therefore protected in its material core by Art. 79.3.
77 See already C. Tomuschat, ‘Die Europäische Union unter der Aufsicht des Bundesverfas-
sungsgerichts’, EuGRZ (1993) p. 489; K.M. Meessen, ‘Maastricht nach Karlsruhe’, NJW (1994)
p. 549, 550 et seq. In the context of the Lisbon judgment, see now R. Bieber, ‘An Association of
Sovereign States’, 5 EuConst (2009) p. 391, 396.
78 These privileged applicants are: the Federal Government, a Land government, or one fourth
(until 30 Nov. 2009 a third) of the members of the Bundestag.
79 Thym, supra n. 38 at p. 1796 et seq.
80 German BVerfG Treaty of Lisbon, supra n. 21 at paras. 172 et seq., in particular 182.
violated, but also that the principle of the social state was encroached upon.81 In
contrast, it considered the complaints to be inadmissible as far as they were based
on an alleged infringement of the rule of law and the separation of powers.82
The second extension of the right of scrutiny relates to the pre-constitutional
(sic!) concept of the constituent power of the people and thus virtually transcends
the legal order of the Basic Law.83 According to the Bundesverfassunsgericht, a
complainant can rely on Article 38 § 1 in order to challenge the loss of sovereign
statehood because the only power with the right to repeal the Basic Law – and
with it the German state – is the constituent power of ‘the people.’84 The key idea
is that what is exclusively reserved to the pouvoir constituant must not be touched
by the pouvoir constitué.85 The court held that the ‘pre-constitutional right’ to give
oneself a constitution86 is not prescribed but merely declaratively mirrored in
Article 146 of the German Basic Law.87 To frame it differently, Article 38 § 1
ensures an inner-systemic right of participation within the existing system, while
Article 146 reflects an outer-systemic right of participation to create a new system.
The logical fracture is that, according to the Bundesverfassungsgericht, the inner-
systemic voter shall be entitled, by relying on Article 38 § 1, to become the guard-
ian of the outer-systemic constituent power reflected in Article 146!88
In brief, the Bundesverfassungsgericht put the right to vote in the centre of its
reasoning and thus enabled several claimants, acting in private capacity, to chal-
lenge the parliamentary approving act as well as the accompanying laws. Ulti-
mately, the German court found the approving act to be compatible with the
constitution. However, it declared the accompanying legislation unconstitutional
to the extent that it did not meet the Court’s demands for an adequate degree of
parliamentary ‘responsibility for integration.’ The particular quirk of this outcome,
unique in Europe, was that the Bundesverfassungsgericht allowed ratification only
8¹ Ibid., paras. 168, 181.

8² The BVerfG took the view that the complainants had not convincingly established a ‘neces-
sary connection’ with the principle of democracy in this regard, ibid., para. 183.
8³ Critically concerning the pre-constitutional construction also M. Jestaedt, ‘Warum in die
Ferne schweifen, wenn der Maßstab liegt so nah?’, 48 Der Staat (2009) p. 496, 501 and 512 et seq.
84 For a critique of the BVerfG’s concept of ‘people’, see T. Eijsbouts, ‘Wir sind das Volk: Notes
about the Notion of “The People” as Occasioned by the Lissabon-Urteil’, 6 EuConst (2010) p. 199
et seq.
85 German BVerfG Treaty of Lisbon, supra n. 21 at para. 179 et seq. and 228.
86 Can this really be a ‘right’ in legal categories?
87 Art. 146 states: This Basic Law, which since the achievement of the unity and freedom of
Germany applies to the entire German people, shall cease to apply on the day on which a constitu-
tion freely adopted by the German people takes effect.
88 See German BVerfG Treaty of Lisbon, supra n. 21 at para. 180. See also the critique of D. Hal-
berstam and C. Möllers, ‘The German Constitutional Court says “Ja zu Deutschland!”’, 10 GLJ
(2009) p. 1241, 1256.
under the condition that a new accompanying legislation fulfilling its demands
came into force. The German legislator followed the court by enacting a new
package of legislation, including the so-called ‘Responsibility for Integration Act’
(RIA).89 After the Bundesverfassungsgericht had declared the remedies directed
against these new laws inadmissible,90 the package could enter into force and thus
paved the way for ratification.
Not every EU citizen has a comparable right to initiate a de facto full-scale
review of constitutionality regarding the national ratification procedure. Even an
institutionally strong constitutional jurisdiction91 is no guarantee that a citizen
acting in private capacity may question the ratification procedure by means of an
individual complaint.
Austria – total inadmissibility

The restrictive approach which was taken by the Austrian Constitutional Court
illustrates this. In 2005 the Austrian Court had already rejected the remedies against
the Constitutional Treaty as inadmissible. In 2008 it proceeded similarly with two
individual petitions (Individualanträge) directed against the ratification of the
Lisbon Treaty. The petitioners argued in particular that the ratification would
amend the basic principles of the Austrian constitutional order in the sense of a
‘total revision’ and therefore require a national referendum.92 By order of 30 Sep-
tember 2008 the Austrian Court denied the admissibility, stating that neither the
act of ratification nor the parliamentary resolution authorising it could be chal-
lenged. Also the Treaty itself could not, according to the Court, be subject to review
as long as it was not in force and therefore not published yet in the Austrian
89 The RIA is contained in Art. 1 of the new Act Extending and Strengthening the Rights of the
Bundestag and the Bundesrat in European Union Matters of 22 Sept. 2009, I no. 60 Official Fed-
eral Law Gazette (2009) p. 3022 et seq. The very same day, two other laws were passed, concerning
the cooperation between Federal Government and Bundestag in EU matters (I no. 60 Official Fed-
eral Law Gazette (2009) p. 3026 et seq.) as well as the cooperation between the Federal state and
the Länder in EU matters (I no. 60 Official Federal Law Gazette (2009) p. 3031 et seq.). For a com-
ment, see M. Nettesheim, ‘Die Integrationsverantwortung – Vorgaben des BVerfG und gesetzgeber-
ische Umsetzung’, 63 NJW (2010) p. 177 et seq.
90 German BVerfG, Case 2 BvR 2136/09 Accompanying Laws to the Lisbon Treaty, order of
22 Sept. 2009.
9¹ For an overview over the ‘constitutional’ jurisdictions in the EU member states, see F.C.
Mayer, ‘Multilevel Constitutional Jurisdiction’, in A. v. Bogdandy and J. Bast (eds.), Principles of
European Constitutional Law, 2nd edn. (Hart 2009) p. 399, 400 et seq.
9² Under Art. 44.3 of the Austrian Federal Constitutional Law any ‘total revision of the Federal
Constitution shall ... be submitted to a referendum by the entire nation, whereas any partial revi-
sion requires this only if one third of the members of the National Council or the Federal Council
so demands.’
official law gazette.93 Hence, in Austria an individual had no legal means of pre-
ventively challenging the Lisbon Treaty under the given procedural circumstances,
while in Germany individual complainants could attack the parliamentary act
approving the Lisbon Treaty before the German ratification was being completed.94
By order of 12 June 2010 the Austrian Constitutional Court also finally re-
jected an individual petition that had been filed after the coming into force of the
Lisbon Treaty. This time the petitioner was a group of deputies who had not
achieved the necessary majority in parliament to imperatively demand an objective
review of constitutionality.95 They therefore filed an individual petition. Under
Austrian constitutional law such a petition is declared admissible only if the peti-
tioner establishes a prima facie infringement of ‘personal rights’ that affects him
or her ‘directly’. The deputies claimed, inter alia, that their constitutional right to
participate in a national referendum was infringed and that the conferral of com-
petences to the European Union limited the constitutionally required contribution
of the deputies to the exercise of legislative powers as guaranteed in Article 24
(legislative power of parliament) and Article 26 § 1 (right to vote) of the Austrian
Federal Constitutional Law. However, the Constitutional Court found that the
petitioners did not sufficiently establish that there was an infringement on per-
sonal rights affecting them directly.96 A comparable criterion of being directly
affected is also required for the admissibility of constitutional complaints under
German constitutional law. But interestingly, in spite of these similarities, the
Austrian Court and the German Bundesverfassungsgericht took a completely dif-
ferent approach.
Latvia – a middle way

The Latvian Constitutional Court went a middle way between the Austrian solu-
tion of total inadmissibility and the German approach of a de facto all-encompass-
ing objective review. Within the framework of a constitutional complaint,97 the
Latvian Court had to decide if the applicants’ fundamental rights under Article
101 of the Latvian constitution were infringed because the ratification of the
Lisbon Treaty had – like in Austria – not been submitted to a national referendum.98
According to Article 101, every Latvian citizen ‘has the right, as provided for by
9³ Austrian CC Treaty of Lisbon I, supra n. 18, point II.2.

94 German BVerfG Treaty of Lisbon, supra n. 21 at para. 170 in line with established case-law.
95 The necessary threshold consists of one third of the National Council’s (chamber of deputies)
members, Art. 140.1 sentence 2 of the Austrian Federal Constitutional Law.
96 Austrian CC Treaty of Lisbon II, supra n. 24, point II.3.
97 Cf. Arts. 16 no. 1, 17.2 no. 11 and 192 (sic) of the Latvian Act on the Constitutional Court.
98 Latvian CC Treaty of Lisbon, supra n. 20.
law, to participate in the work of the State and of local government, and to hold
a position in the civil service. (...)’
In its judgment of 7 April 2009, the Latvian Court declared the case admissible
on the basis of an alleged violation of Article 101 which, according to the Court,
protects the individual right to participate in a referendum as far as the latter is
provided for in the constitution or in another normative act.99 Two constitu-
tional provisions were claimed to require a national referendum in the present
context. The first provision was Article 68.4, according to which ‘substantial
changes’ regarding EU membership shall be decided by a national referendum if
at least one-half of the members of parliament so request. The second stipulation
was Article 77, which requires a referendum for the amendment of certain funda-
mental articles, such as the sovereignty-clause in Article 2.100 As the Constitu-
tional Court declared the case admissible on the basis of the alleged violation of
Article 101 in connection with these two provisions, it consequently limited its
scrutiny to the question of whether the Lisbon Treaty had been ratified in compli-
ance with the procedures established in the constitution.101 Article 68.4 did not
provoke major problems in this respect. Its wording alone indicates that the ques-
tion whether to submit ratification to a referendum or not lies within the sole
hands of parliament.102 In contrast, the question of whether the ratification of the
Lisbon Treaty touches upon the principle of sovereignty as enshrined in Article 2
and therefore requires a referendum under Article 77, required the Latvian Court
to present more detailed reasoning to demonstrate that it did not.103
Differences of institutional self-conception?

To sum up, while in France, the Czech Republic, Hungary and Poland the control
of constitutionality was carried out under the procedural rules of objective – ex
ante or ex post – treaty reviews, the constitutional courts in Germany, Austria and
Latvia were confronted with individual complaints or petitions.
Although the Austrian, German and Latvian constitutional courts are all insti-
tutionally ‘strong’ constitutional courts in the classic sense, they came to signifi-
cantly differing solutions concerning the question of admissibility. While the
Austrian Constitutional Court rejected the complaints as entirely inadmissible,
the German Bundesverfassungsgericht carried out a de facto full-scale review of
99 Ibid., points 9 and 13.

¹00 Art. 2 reads as follows: ‘The sovereign power of the State of Latvia is vested in the people of
Latvia.’
¹0¹ Latvian CC Treaty of Lisbon, supra n. 20, points 9 et seq.
¹0² Ibid., point 19.4.
¹0³ Ibid., points 16.1–18.10. Thus the Latvian CC also rejected a violation of Art. 101.
constitutionality. The Latvian Constitutional Court in turn went a middle way as

it affirmed admissibility but limited its scrutiny to specified constitutional grounds
for review.
Taking into consideration that the relevant procedural provisions in Austria
and Germany share significant similarities, one might ask if the opposing results
can be explained by different forms of institutional self-conception or varying
degrees of judicial self-confidence. While the Austrian Court sticks closely to the
wording of the procedural provisions and thus precludes a private person from
preventively challenging the treaty reform, the German Bundesverfassungsgericht
creates a de facto preventive treaty review by means of judicial interpretation. Do
not be mistaken: The decision of the French Conseil constitutionnel and the obiter
dictum of the Hungarian Constitutional Court demonstrate clearly how practical
and effective an ex ante treaty review can turn out to be. But the crucial question
remains if it is up to a judicial body to create such a procedure if is not foreseen
by the constitution or when the procedural conditions for an objective review of
constitutionality are not met in the particular case. The approach of the Bundes-
verfassungsgericht is even more dubitable if one takes into consideration that this
court’s former competence to deliver advisory opinions was abolished in the early
years of the Basic Law. Seen in this light, the Bundesverfassungsgericht’s approach of
declaring the case admissibility comes close to an act of constitutional ‘self-
authorisation’.104
Legitimizing the EU: A case of multi-level democracy?

As all decisions under review paved the way for ratification in one way or another,
the substantial key question is their legal impact on the future development of
European law and policies. The Lisbon decisions underlined that a decisive factor
in this respect is the courts’ visions on democratic legitimation of EU public au-
thority. On this point the Lisbon decisions reveal considerable differences.
Parliamentary assent to the future application of ‘dynamic treaty provisions’

The different visions of the national courts on the democratic legitimation of EU
public authority become particularly apparent when it comes to the question
whether and if so to what extent prior assent of national parliaments is necessary
in cases when EU (treaty-) law is developed dynamically, i.e., without an ordinary
amendment procedure.
¹04 For judicial ‘acts of self-authorisation’, see – in a historical-comparative perspective –

D. Herrmann, ‘Akte der Selbstautorisierung als Grundstock institutioneller Macht von Verfas-
sungsgerichten’, in H. Vorländer (ed.), Die Deutungsmacht der Verfassungsgerichtsbarkeit (VS-Verlag
2006) p. 141, 157 et seq.
Five categories of so-called105 ‘dynamic treaty provisions’ can be distinguished.

The first is the simplified treaty revision procedure according to the general clause
in Article 48.6 TEU and the specific provisions in Article 42.2 (1) TEU and Ar-
ticles 25.2, 218.8 (2), 223.1 (2), 262, 311.3 TFEU. The second group is composed
of the ‘bridge’ or ‘passerelle’ mechanisms enshrined in the general clause of Article
48.7 TEU106 and the specific regimes under Article 31.3 TEU and Articles 81.3
(3), 153.2 (4), 192.2 (2), 312.2 (2), 333.1, 333.2 TFEU. The third category is
the flexibility clause in Article 352 TFEU (ex-Article 308 TEC). The fourth case
concerns the so-called ‘emergency brakes’ under Articles 48.2, 82.3 and 83.3
TFEU, which allow for the preliminary suspension of the ordinary legislative
procedure if a member state so requests. Finally, the fifth category relates to spe-
cific stipulations according to which the Council – after obtaining the consent of
the European Parliament – can adopt unanimously decisions in ‘sensitive’ fields
such as criminal law (Article 83.1 (3) TFEU). It is important to note that only the
first and the second category concern the simplified (and insofar ‘dynamic’) amend-
ment of EU treaty law. The other categories relate essentially to the legislative
process at EU level. Taking into account the demanding procedural requirements
of categories 3-5, one may doubt if the term ‘dynamic treaty provisions’ as used
by the Bundesverfassungsgericht107 is an adequate description at all.
Similarities and peculiarities

The Bundesverfassungsgericht is the only constitutional court in Europe that demands
a constitutive authorisation of national parliament in all five cases,108 be it by act
of parliament or by parliamentary resolution.109
As far as the simplified revision procedure under Article 48.6 is concerned, the
German court takes a similar approach as the French Conseil constitutionnel, which
ruled in both decisions that the ratification of a simplified revision under Article
48.6 TEU requires the authorisation of the French parliament.110 In this respect
the German court refers explicitly to its French counterpart.111
¹05 Cf. German BVerfG Treaty of Lisbon, supra n. 21 at para. 239.

¹06 Cases in which Art. 48.7 could be applied are, for instance, Arts. 82.2 (2) (d), 83.1 (3), 86.4
and 308.3 TFEU.
¹07 Supra n. 105.
¹08 See German BVerfG Treaty of Lisbon, supra n. 21 at paras. 412-419.
¹09 For the differences, see in detail I. Pernice, ‘Motor or Brake for European Policies? Germany’s
New Role in the EU after the Lisbon Judgment of Its Federal Constitutional Court’, in I. Pernice
and J.M. Beneyto Pérez (eds.), Europe’s Constitutional Challenges in the Light of the Recent Case Law:
Lisbon and Beyond (Nomos 2011 forthcoming), point IV.
¹¹0 French CC Constitutional Treaty, supra n. 12 at para. 36 and later French CC Treaty of Lis-
bon, supra n. 17, para. 26. According to the CC Art. 53 of the French Constitution applies in this
respect.
¹¹¹ German BVerfG Treaty of Lisbon, supra n. 21 at para. 312 with reference to French CC
Treaty of Lisbon, supra n. 17 at para. 26.
But here the parallels between the two judicial bodies end.
As to the application of bridge-clauses, the French Conseil constitutionnel does
not demand a prior assent by the French parliament. Instead these clauses are one
reason for the Conseil constitutionnel to demand a constitutional amendment for
the ratification of the Lisbon Treaty.112 The constitutional authorisation which has
been introduced in the French constitution in the aftermath of the Lisbon decision
(former Article 88-1 § 2) and which allowed the ratification of the Lisbon Treaty
thus already encompassed and insofar anticipated all future applications of pas-
serelle-clauses. Therefore the French representative in the Council is entitled to
vote for such an application without being previously authorised by parliament
to do so. This entails important consequences for the judicial branch as well. As
there is no requirement of prior parliamentary assent, the future application of
passerelle-clauses cannot be subjected to constitutional review in France. In con-
trast, in Germany each application of a passerelle-clause can be challenged before
the Bundesverfassungsgericht, simply because the authorising parliamentary act can
be subjected to constitutional review.
According to the German court, the foundation of the requirement of the
parliament’s ‘positive’ involvement is parliament’s ‘responsibility for integration’
– which is in fact also a judicial responsibility.113 This applies to all forms of dy-
namic forms of development mentioned above.114
National v. multi-levelled democracy

The peculiarities of the stance of the Bundesverfassungsgericht become even more
apparent if it is contrasted to the Lisbon judgments of the Czech, the Latvian and
the Hungarian Constitutional Courts and the Polish Constitutional Tribunal. Like
the French Conseil constitutionnel, these courts did not make the ratification of the
Lisbon Treaty conditional on the passage of accompanying legislation ensuring
the prior assent of national parliaments in the above-mentioned cases. Instead,
they highlight that the parliamentary right to veto is a sufficient procedural safe-
guard.115 The Latvian CC ruled that
The TL would introduce additional democratic guarantees, namely, national Parlia-

ments would have the right to object if the EU, based on Article 352 of the TFEU,
¹¹² French CC Constitutional Treaty, supra n. 12 at paras. 33-35 and French ConC Treaty of
Lisbon, supra n. 17, paras. 23 et seq. and 27.
¹¹³ See explicitly German BVerfG Treaty of Lisbon, supra n. 21 at para. 236 at the end.
¹¹4 In contrast, particularly Art. 352 TFEU does not seem to be a constitutional problem for the
French CC at all.
¹¹5 See Czech CC Treaty of Lisbon I, supra n. 19 at paras. 161-164, 172-175; Czech CC Treaty
of Lisbon II, supra n. 22 at para. 134; Latvian CC Treaty of Lisbon, supra n. 20, point 18.6; Hungar-
ian CC Treaty of Lisbon, supra n. 23, point IV.2.5 and Polish CT Treaty of Lisbon, supra n. 25.
will draft new legal acts. ... Accordingly, the Constitutional Court concludes that
Latvia will have the rights and the ability to block changes in the decision-making
procedure that are undesirable for Latvia and the Saeima will have the possibility to
express its opinion before changes come into force.116
The Czech Constitutional Court additionally draws on the jurisprudence of the

Court of Justice in order to establish that Article 352 TFEU is not a ‘blanket norm’
for extending the competences of the Union.117 Further, it emphasises the decisive
role of the European Parliament, whose consent is not only mandatory under
Article 352.1 TFEU but also under Article 48.7(4) TEU. It is precisely in this
context that the Czech court in its first Lisbon decision underlines the multi-levelled
nature of the democratic legitimation of EU authority.118 In their second decision,
the Czech judges highlight this point even further. They quote the opinion of
former Advocate General Miguel Poiares Maduro regarding the multi-levelled
character of the principle of representative democracy119 and openly object to the
argument of the Bundesverfassunggericht:
Insofar as [Article 10.1] of the TEU provides that “The functioning of the Union
shall be founded on representative democracy”, that does not mean that only proc-
esses at the European level should ensure fulfilment of that principle. That article is
directed at processes both on the European and the domestic level, not only at the Euro-
pean Parliament, as stated by the German Constitutional Court in point 280 of its deci-
sion (...).
In other words, the democratic process on the Union and domestic levels mutually
supplement and are dependent on each other. (...)
For similar reasons, one cannot see conflict of Article 14.2 of the TEU, which gov-
erns the number of members of the European Parliament, with the principle of
equality (...). As pointed out above, the European Parliament is not the exclusive
source of democratic legitimacy for decisions adopted on the level of the European
Union. That is derived from a combination of structures existing both on the domestic
and on the European level, and one cannot insist on a requirement of absolute equality
among voters in the individual Member States.120
¹¹6 Latvian CC Treaty of Lisbon, supra n. 20, point 18.6 (emphasis added).
¹¹7 Czech CC Treaty of Lisbon I, supra n. 19 at para. 151 et seq.
¹¹8 Ibid., para. 173: ‘The Treaty of Lisbon transfers powers to bodies that have their own regu-
larly reviewed legitimacy, arising from general elections in the individual member states. Moreover,
the Treaty of Lisbon permits several ways of involving domestic parliaments (the possibility for a
parliament, or one of its chambers, to directly express its lack of consent, is one of the forms of participa-
tion by domestic parliaments)’ (emphasis added).
¹¹9 Czech CC Treaty of Lisbon II, supra n. 22 at para. 138 with reference to the opinion of former
A.G. Poiares Maduro of 26 March 2009, Case C-411/06, Commission v. Parliament and Council.
¹²0 Ibid., para. 137 et seq. (emphasis added).
The Czech Constitutional Court thus puts a certain degree of trust in the multi-
levelled structure of democracy in the European Union and particularly in the
ability of the European Parliament to provide for a genuine – albeit not exclusive
– link of democratic legitimacy, even despite the EP’s degressively proportional
composition. In addition, the Czech court highlights the concept of ‘pooled’
sovereignty.121
In contrast, the Bundesverfassungsgericht takes the view that the necessary degree
of democratic legitimacy of EU public authority can – at the moment – only
derive from the national ‘state people’ (Staatsvolk). As a consequence, the German
court needs to tie the application of the bridging-clauses and the other provisions
mentioned above to the prior and constitutive assent of the German parliament.
The autonomous democratic mechanisms and institutions on EU-level have, so
the argument goes, a complementary character at best, but not a constitutive one.
In so far as the people itself is not directly called upon to decide, democratic legitima-
tion can only be achieved by means of parliamentary responsibility. (...) In so far as the
Member States elaborate treaty law in such a way as to allow treaty amendment
without a ratification procedure solely or mainly by the institutions of the Union,
albeit under the requirement of unanimity, whilst preserving the principle of confer-
ral, a special responsibility is incumbent on the legislative bodies, in addition to the
Federal Government, within the context of participation which in Germany, has to
comply internally with the requirements under Article 23.1 of the Basic Law (re-
sponsibility for integration) and which may be invoked in any proceedings before the
Federal Constitutional Court. (...)
Measured against requirements in a constitutional state, even after the entry into
force of the Treaty of Lisbon, the European Union lacks a political decision-making
body created in equal elections by all citizens of the Union and with the ability to
uniformly represent the will of the people. (...) Even in the new wording of Article
14.2 Lisbon TEU, and contrary to the claim that Article 10.1 Lisbon TEU seems
to make according to its wording, the European Parliament is not a representative body
of a sovereign European people. This is reflected in the fact that it is designed as a
representation of peoples in the respective national contingents of Members, not as
a representation of Union citizens in unity without differentiation, according to the
principle of electoral equality. (...)
The deficit of European public authority that exists when measured against requirements
on democracy in states cannot be compensated for by other provisions of the Treaty of
Lisbon and, to that extent, it cannot be justified. (...).122
If one was asked to trace back the arguments of the Bundesverfassungsgericht to a

leading principle, it would neither be sovereignty nor identity, but democracy. The
¹²¹ Ibid., para. 147 and Czech CC Treaty of Lisbon I, supra n. 19 at para. 104.
¹²² German BVerfG Treaty of Lisbon, supra n. 21 at paras. 236, 280 and 293 (emphasis added).
principle of democracy is the corner stone of the German Lisbon decision. It is the
substantial key argument for the German court to declare the case admissible on
the basis of an alleged infringement of the right to vote. Moreover, the principle
of democracy underlies and even predetermines the principle of sovereign state-
hood as developed in the Lisbon judgment.123 But above all, the principle of de-
mocracy is placed at the heart of Germany’s constitutional identity, for the right
to free and equal participation in public authority is, according to the Bundesver-
fassungsgericht, ‘enshrined in human dignity’ itself.124
The tragedy is that the Bundesverfassungsgericht’s conception of democracy is
existentially bound to the (pre-)existence of statehood and is blind to constitutive
forms of democratic legitimation within multi-levelled entities. The court thus
leaves the Germans with an astonishing binary choice: They may either remain
part of the so-called association of sovereign and ‘fully democratically’ (volldemo-
kratisch) organised states whose ‘peoples’ remain the only subjects of democratic
legitimation. Or they may participate in the creation of a European federal state,
which would require a change of the ‘subject of democratic legitimation’125 and
the superseding of one of Germany’s most vaunted post-war-inventions: the Basic
Law.126
If the Germans do not wish to give up their constitution – and who would be
surprised if they didn’t want to? –, the application of all forms of ‘dynamic treaty
provisions’ must be previously legitimised by the German parliament and remain
under the control of the Bundesverfassungsgericht. To frame it differently, as long
as the German ‘people’ does not opt for a European federal state, dynamic treaty
mechanisms shall not be too dynamic. Here it becomes clear that the specificities
of the Bundesverfassungsgericht´s conception of democracy entail direct conse-
quences for the future of European constitutional development.
Who demands prior parliamentary assent: division of powers I

One might object that the strong critique which is directed against the Bundesver-
fassungsgericht’s Lisbon decision inside and outside Germany,127 ignores that Ger-
many is by far not the only country in which prior parliamentary assent is deemed
¹²³ Ibid., para. 248: ‘The safeguarding of sovereignty, demanded by the principle of democracy in
the valid constitutional system ...’ (emphasis added).
¹²4 Ibid., para. 211.
¹²5 The BVerfG demands that democratic requirements in this new entity would have ‘to be
fully consistent with the requirements for the democratic legitimation of a union of rule organised
by a state’, cf. ibid., para. 263. However, it is not clear what the normative basis for this claim shall
be, as it can’t be the (then superseded) Basic Law.
¹²6 Ibid., paras. 179, 229, 263, 298 and 334. See also D. Halberstam and C. Möllers, supra
n. 88 at p. 1255 et seq.
¹²7 Cf. supra n. 21.
necessary for the application of dynamic treaty provisions. Examples for such a
requirement can also be found in section 6 of the British European Union Amend-
ment Act of 2008128 (to which the German court refers129) and in the new Article
23i of the Austrian Federal Constitutional Law, which even requires a qualified
majority in both houses of parliament. Also, the Czech legislator in 2009 passed
a statute establishing such requirements, not only with respect to the passerelle-
clauses but also with regard to Article 352 TFEU.130
So, much ado about nothing? Not really, because even if such requirements can
be found in several member states, the question still remains who took the decision
to initiate them. This leads to a second peculiarity. While in Germany it was a
court which forced the legislator to establish precisely defined rules to exercise
parliamentary responsibility for integration, in all other countries such a decision
was taken freely by the (constitutional) legislator. In particular the Czech Consti-
tutional Court left it to the legislative branch to decide how and to what extent
national parliaments’ rights are to be framed:
However, in this regard we cannot help but see that there are as yet no related provi-
sions in the legal order of the Czech Republic that would allow implementation of
the decision making procedures set forth in paragraphs six and seven of Art. 48 on
the domestic level. The absence of these procedures, in and of itself, does not affect
the question of whether the Treaty of Lisbon is constitutional, but because the Treaty of
Lisbon presumes the intervention of domestic parliaments, the government, as the
sponsor of the Treaty of Lisbon (...) should reflect that in a timely manner and ade-
quately, by proposing relevant procedures on the domestic level, and should ensure that
the Treaty is compatible and interconnected with the constitutional order of the Czech
Republic, not only in view of the participation of the parliament, but also in view of
the possibility of preliminary review of an amendment of the Treaties by the Con-
stitutional Court.131
In other words, the comparative perspective reveals how deeply the German court
mistrusts not only the EU political process, but also the political and parliamen-
tary process in Germany.132
¹²8 British European Union (Amendment) Act of 19 June 2008, available at <www.statutelaw.
gov.uk/content.aspx?activeTextDocId=3490003>.
¹²9 German BVerfG Treaty of Lisbon, supra n. 21, at para. 320.
¹³0 Act No. 162/2009.
¹³¹ Czech CC Treaty of Lisbon I, supra n. 19 at para. 165, confirmed in Czech CC Treaty of
Lisbon II, supra n. 22 at para. 134.
¹³² This is a line of argument which also characterised the European Arrest warrant case. It is
probably one of the great ironies that it is a constitutional court which obliges the national parlia-
ment to observe its responsibility for integration.
Popular vote as a requirement for ratification?

The question of democratic legitimation of EU public authority is of course not
limited to the roles of European and national parliaments. It also concerns forms
of direct democratic expression.
As seen above, both the Austrian as well as the Latvian Constitutional Courts
were being confronted with the question whether the ratification of the Lisbon
Treaty required a national referendum or not.133 The Latvian court held that a
referendum was not mandatory,134 while the Austrian court declared the indi-
vidual petitions inadmissible, laconically repeating its established jurisprudence
according to which an individual has only a right to participate in a legally arranged
referendum, but not the right to demand such a referendum.135 The Bundesverfas-
sungsgericht in turn stated, for the first time, that a (hypothetical) participation of
Germany in a federal European state would require a new constitution by refer-
endum under the pre-constitutional right reflected in Article 146.
The only country in which a referendum was held on the Lisbon Treaty was
Ireland.136 There, the Crotty decision of the Irish Supreme Court (dating back to
1987)137 initiated a constitutional practice according to which the Irish constitu-
tion is supplemented with a specific authorisation for ratification each time a
major treaty reform is under way.138 As an amendment of the Irish constitution
requires a national referendum,139 the ratification of every major treaty reform is
consequently presumed by a popular vote.140
In all other member states a referendum was not deemed necessary, particu-
larly not in France and the Netherlands. In France the decision whether or not to
hold a referendum is according to Article 11 of the French constitution a discre-
¹³³ See also the High Court for England and Wales (Queen’s Bench Division, Divisional Court),
R (on the application of Stuart Wheeler) v. Prime Minister and Foreign Secretary [2008] EWHC 1409
(Admin), which ‘found nothing in the claimant’s case to cast doubt on the lawfulness of ratifying
the Lisbon Treaty without a referendum’ (para. 59).
¹³4 Latvian CC Treaty of Lisbon, supra n. 20, points 18 and 19.
¹³5 Austrian CC Treaty of Lisbon II, supra n. 24, point II.3.1.
¹³6 The first referendum was held on 12 June 2008 and resulted in a negative outcome. The
second referendum was held on 2 October 2009 and resulted in a positive outcome. See on that
M. Cahill, ‘Ireland’s Constitutional Amendability and Europe’s Constitutional Ambition: the Lis-
bon Referendum in Context’, 9 GLJ (2008) p. 1191 et seq.
¹³7 Irish Supreme Court, Case 1986 No. 12036P, Crotty v. An Taoiseach [1987] IR 713.
¹³8 Now Art. 29.4 No. 5 of the Irish constitution. As to the Constitutional Treaty, see G Hogan,
‘Ratification of the European Constitution – Implications for Ireland’, in A. Albi and J. Ziller
(eds.), The European Constitution and National Constitutions (Kluwer 2007) p. 137 et seq.
¹³9 Art. 46.2 of the Irish constitution.
¹40 Whether an express constitutional authorisation is really necessary under the Crotty-doc-
trine is contentious, see, e.g., A. Maurer and B. Roth, ‘Warum Irland abstimmen muss(te) – oder
auch nicht’, available at <www.swp-berlin.org/common/get_document.php?asset_id=5575>.
tionary power of the President of the Republic. Unlike his predecessor Jacques
Chirac, who put the Constitutional Treaty to a popular vote,141 Nicolas Sarkozy
opted against a referendum in the case of the Lisbon Treaty.
In the Netherlands the referendum on the Constitutional Treaty was legally
speaking purely consultative.142 In its opinion on the mandate of the IGC in 2007,
the Dutch Council of State underlined that regarding the Lisbon Treaty a con-
sultative referendum was constitutionally admissible, but not required, neither by
constitution nor by the mere precedent of the Constitutional Treaty referendum.143
In Denmark the decision not to hold on the Lisbon Treaty was preceded by an
astonishing opinion of the Ministry of Justice. According to the wording of Arti-
cle 20 of the Danish constitution, which allows the ‘delegation’ of competences,
a referendum is required only if the specific majority of fifth-sixths of the members
of parliament is not achieved. However, in Denmark a constitutional practice has
evolved, according to which a referendum is generally held when Article 20 applies,
regardless of whether the qualified majority has been met or not.144 The Danish
Ministry of Justice took the view that, in contrast to the Constitutional Treaty, the
Lisbon Treaty would not entail a ‘delegation’ of competences in the sense of Arti-
cle 20 and thus would not require a (customary) referendum.145 This legal evalu-
ation is neither convincing nor consistent with the Ministry’s previous evaluation
¹4¹ The referendum concerning the Constitutional Treaty was appointed by presidential Decree
No. 2005-218 of 9 March 2005.
¹4² See L.F.M. Besselink, ‘Constitutional Referenda in the Netherlands: A Debate in the Mar-
gin’, 11 EJCL (May 2007) p. 1, 14 and id, ‘The Dutch Constitution, the European Constitution
and the Referendum in the Netherlands’, in A. Albi and J. Ziller (eds.), The European Constitution
and National Constitutions (Kluwer 2007) p. 113, 118.
¹4³ Dutch State Council Lisbon-Mandate, supra n. 27, point 4.2: ‘There is nothing in the Con-
stitution to prevent the legislator from holding such referendums on an ad hoc basis, provided that
it indicates that there is a special justification for departing from the self-contained arrangements
for approving treaties and provided that the referendum procedure is governed by an Act of Parlia-
ment. Mere precedent will not suffice. That would create a substantive basis for the referendum as
a structural instrument (in this case, for use when approving treaties) which would not be in keep-
ing with the self-contained arrangements in the Constitution.’ See on that J. Ziller, ‘The Law and
Politics of the Ratification of the Lisbon Treaty’, in S. Griller and J. Ziller (eds.), The Lisbon Treaty
(Springer 2008) p. 309, 317 et seq.
¹44 See H. Koch, ‘The Danish Constitutional Order’, in A.E. Kellermann et al. (ed.), EU-En-
largement – The Constitutional Impact at EU and National Level (TMC Asser Press 2001) p. 109,
111. Critically to this practise H. Rasmussen, ‘Denmark’s Waning Constitutionalism and Article 20
of the Constitution on Transfer of Sovereignty’, in A. Albi and J. Ziller (eds.), The European Consti-
tution and National Constitutions (Kluwer 2007) p. 149, 150.
¹45 Opinion of 4 Dec. 2007. In the aftermath of this opinion, on 11 Dec. 2007 the Danish
government took the decision to ratify the Lisbon Treaty according to the ordinary procedure under
Art. 19 of the Danish constitution. The approval of the Danish parliament followed on 24 April
2008.
of the Constitutional Treaty. It is clear that the political desire to avoid holding a
referendum determined the legal argument.
In summary, the ratification of the Lisbon Treaty shows that in almost all
member states a popular vote is not viewed as being constitutionally mandatory
for ratification, even in cases of fundamental treaty reform. Instead, in most con-
stitutional orders that allow a consultative or legally binding referendum, the final
decision as to whether to hold a referendum rests with the legislative or the ex-
ecutive branch.146 With the exception of Ireland, all countries in which a referen-
dum would have been constitutionally possible avoided this for obvious reasons.
Constitutional limits and judicial reservations

The supreme jurisdictions also shape future developments of European constitu-
tionalism by establishing constitutional limits and judicial reservations. Here we
touch upon the legal consequences of the courts’ conceptions of sovereignty and
(national) constitutional identity.
Substantive constitutional limits to future developments of EU law

The first point in this regard aims at substantive constitutional limits to the future
development of European Union primary law.
Two categories of constitutional limits

It is imperative to distinguish two categories of constitutional limits in this respect.
On the one hand there are limits that indicate the necessity of a constitutional
amendment, on the other hand there are ‘red lines’ marking the inalienable sub-
stantive core of a constitutional order which, in some countries, is not even subject
to constitutional revision.
The jurisprudence of the French Conseil constitutionnel provides an example of
the first category. According to established case-law,147 the Conseil verifies wheth-
er the ratification of a new treaty requires a prior constitutional amendment.
According to the Conseil, a revision of the French constitution is necessary if the
treaty in question contains a clause expressly contrary to the French constitution
¹46 See particularly Arts. 3a.2 of the Slovenian, 68.4 of the Latvian, 90.3 of the Polish and 10a.2
of the Czech constitution, allowing the national parliament to take a decision to substitute or com-
plement its own act of approval by a referendum. In France, according to Art. 88-5 of the constitu-
tion any bill authorizing the ratification of a treaty pertaining to the accession of a state to the EU
shall be submitted to referendum by the President of the Republic. But also here, by passing a mo-
tion adopted in identical terms in each house by a three-fifths majority, parliament may authorize
the passing of a bill according to the (parliamentary) constitutional amendment procedure.
¹47 See in particular French CC, Case No. 92-308 DC Maastricht I, decision of 9 April 1992.
(1), calls into question constitutionally guaranteed rights and freedoms (2) or af-
fects the ‘essential conditions of the exercise of national sovereignty’ (3).148 The
latter is the most important criterion in practice, as the judgments on the Consti-
tutional Treaty and the Lisbon Treaty show. Both treaties jeopardised these ‘es-
sential conditions’ and thus required a constitutional amendment. Although
Article 89.5 of the French constitution sets potential limits to constitutional
amendments,149 the Conseil constitutionnel has, until today, not relied on this
provision in order to establish constitutional limits to European integration. Hence,
Article 89.5 does not belong to the relevant norms of reference, neither in the
decision on the Constitutional Treaty nor in that on the Lisbon Treaty. Moreover,
the Conseil takes the general view that it is not authorised to review constitu-
tional amendments.150 The Conseil constitutionnel thus has been described aptly
as a pointsman (aiguilleur) which only indicates whether ratification requires
amendment of the constitution or not.151 A further example of the first category
is that a constitutional court accepts constitutional limits to the national integra-
¹48 Cf. French CC Constitutional Treaty, supra n. 12 at para. 7 and Treaty of Lisbon, supra n. 17,
para. 9.
¹49 According to Art. 89.5 of the French constitution, the ‘republican form of government shall
not be the object of any amendment’.
¹50 French CC, Case 2003-469 Decentralisation, decision of 26 March 2003, paras. 2 and 3.
That the Conseil constitutionnel shows extreme reluctance with regard to Art. 89.5 of the French
constitution is also illustrated by its Maastricht decisions. After the Conseil had ruled in a first deci-
sion that the ratification of the Maastricht Treaty required a prior constitutional amendment in
1992 (Case 92-308 DC Maastricht I, decision of 9 April 1992) it was asked to review the constitu-
tionality of the Maastricht Treaty a second time. The group of senators who introduced the second
proceeding was unsatisfied that the French constituent authority had not amended pre-existing
provisions, such as the sovereignty-clause under Art. 3, but had decided merely to supplement the
French constitution with several EU related provisions. In other words, the second proceeding
would have given the Conseil the opportunity to make its views on the constitutionality of the
constitutional amendment. However, the Conseil did not follow the argument of the applicants,
according to which the ratification of the Maastricht Treaty was still unconstitutional. Instead it
decided that ‘subject to the provisions governing the periods in which the Constitution cannot be
revised (Arts. 7 and 16 and the fourth paragraph of Art. 89) and to compliance with the fifth para-
graph of Art. 89 (‘The republican form of government shall not be the object of an amendment’),
the constituent authority is sovereign; it has the power to repeal, amend or amplify constitutional provi-
sions in such manner as it sees fit; there is accordingly no objection to insertion in the Constitution of new
provisions which derogate from a constitutional rule or principle; the derogation may be express or im-
plied’ (French CC, Case 92-312 DC Maastricht II, decision of 2 Sept. 1992, para. 19, emphasis
added). See on that J. Ziller, ‘Sovereignty in France: Getting Rid of the Mal de Bodin’, in
N. Walker (ed.), Sovereignty in Transition (Hart 2003) p. 261, 271 et seq. In its third Maastricht
decision the CC declared itself not competent to decide on the constitutionality of the law approv-
ing the ratification as this law had been subjected to a referendum and was thus a direct expression
of national sovereignty (‘expression directe de la souveraineté nationale’), French CC, Case 92-313
DC Maastricht III, decision of 23 Sept. 1992, para. 2.
¹5¹ L. Favoreu, La politique saisie par le droit (Economica 1988) p. 30.
tion clause, i.e., the clause allowing the conferral of competencies to the Euro-
pean Union, as long as such limits are not constitutionally inalienable. Examples
for such an approach are given by the Lisbon decisions of the Hungarian Consti-
tutional Court152 and the Polish Constitutional Tribunal153 as regards the integra-
tion clauses in both countries.
The most explicit example for the second category is given by the Lisbon judg-
ment of the Bundesverfassungsgericht . As already pointed out, the key argument
of this decision is the principle of democracy as protected in its essential content
by the so-called ‘eternity clause’ under Article 79.3 of the German Basic Law.154
Hence, the absolute constitutional limits to integration under the German Basic
Law begin particularly where the principle of democracy (and with it the right to
vote) would be eroded at its substantial core. In the Lisbon judgment the German
court identified five key areas within which the future conferral of competencies
to the EU would bear a high risk of violating this material core as protected by
the eternity clause:
Particularly sensitive for the ability of a constitutional state to democratically shape

itself are [since time immemorial155] decisions on substantive and formal criminal
law (1), on the disposition of the monopoly on the use of force by the police within
the state and by the military towards the exterior (2), fundamental fiscal decisions
on public revenue and public expenditure, the latter being particularly motivated,
inter alia, by social policy considerations (3), decisions on the shaping of living
conditions in a social state (4) and decisions of particular cultural importance, for
example on family law, the school and education system and on dealing with religious
communities (5).156
This catalogue is spelled out in more detail in the following paragraphs of the
Lisbon judgment.157 In a remarkably apodictic way – which has been vividly
criticised158 – the Bundesverfassungsgericht places constitutional stop signs before
¹5² Hungarian CC Treaty of Lisbon, supra n. 23, point IV.2.3.2.

¹5³ According to the Polish CT Treaty of Lisbon, supra n. 25, this limit is constituted by the key
principles determining the ‘constitutional identity’ of Poland, i.e., the protection of human dignity
and the constitutional rights and freedoms, the respect of sovereign statehood and the principles of
democracy, the rule of law, social justice and the bases of the economic system.
¹54 According to Art. 79.3 of the German Basic Law amendments ‘affecting the division of the
Federation into Länder, their participation on principle in the legislative process, or the principles
laid down in Articles 1 [human dignity – human rights – legally binding force of basic rights] and
20 [constitutional core principles] shall be inadmissible.’
¹55 The little, albeit characteristic part ‘seit jeher’, which suggests historic continuity instead of
providing substantial arguments, was left out in the official English translation.
¹56 German BVerfG Treaty of Lisbon, supra n. 21 at para. 252.
¹57 Ibid., paras. 253-260.
¹58 Vf. C. Schönberger, ‘Lisbon in Karlsruhe: Maastricht’s Epigones at Sea’, 10 GLJ (2009)
p. 1201, 1208 et seq. who speaks even of an ‘arrogation of power’.
further conferrals of competences. We learn, for instance, that the task of securing
the individual’s livelihood is and must remain a ‘primary task’ of the member
states.159 The only reason why the right to mint and issue coins – which has also
been a classical state competence – does not figure in the court’s list is probably
the mere fact that this marque de souveraineté (Bodin)160 has already been conferred
to Union level.161
Who determines constitutional limits: division of powers II

From a comparative perspective, it is not striking that a constitutional court derives
constitutional limits to European integration from an eternity-clause. But the
German Lisbon judgment is exceptional because of the extent to which the Bun-
desverfassungsgericht relies on Article 79.3.162 No other court has spelled out an
eternity clause in such a detailed, albeit apodictic manner as the German court
did with Article 79.3.
As already mentioned, the French Conseil constitutionnel shows extreme reluc-
tance with regard to Article 89.5 of the French constitution as it acts as a proce-
dural pointsman, leaving the question of whether and of how to amend the
constitution to the constitutional legislator.163 A restrictive approach towards
(potential) constitutional eternity clauses can be observed also for Article 139 of
the Italian,164 Article 110 of the Greek,165 Article 288 of the Portuguese,166 Article
152.1 and 152.2 of the Romanian167 and Article 182 of the Cypriot constitution.168
But the peculiarity of the Bundesverfassungsgericht’s approach towards the eter-
nity clause – which was drafted primarily in order to prevent a slide back into

¹60 Vf. J. Bodin, Les six livres de la République, 10th edn. (1593 – reprint 1986), book I, chs. XX
295, 306 and 309.
¹6¹ Similarly C. Schönberger, supra n. 155 at p. 1209.
¹6² Cf. Weber, supra n. 30, who draws a comparable conclusion.
¹6³ See, above all, French CC Decentralisation, supra n. 150.
¹64 M. Cartabia, ‘The Legacy of Sovereignty in Italian Constitutional Debate’, in N. Walker
(ed.), Sovereignty in Transition (Hart 2003) p. 305, 316.
¹65 See J. Iliopoulos-Strangas, ‘Offene Staatlichkeit’, in A. v. Bogdandy and P.M. Huber (eds.)
Ius Publicum Europaeum, vol. II (C.F. Müller 2008) § 16 Griechenland, para. 43 et seq.
¹66 M. Poiares Maduro, ‘EU Law and National Constitutions: Portugal’, manuscript for BIICL
(ed.) FIDE XX Congress London, vol. 1 (BIICL 2002), formerly available at <www.fide2002.org>,
p. 13 et seq.
¹67 V. Duculescu and A. Ruxandra, ‘Romania’, in A.E. Kellermann et al. (eds.), The Impact of
EU Accession on the Legal Orders of New Member States and (Pre-)Candidate Countries (T.M.C.
Asser Press 2006) p. 113, 118 et seq.
¹68 For the legal situation in Cyprus, see N. Emiliou, ‘Cyprus’, in A.E. Kellermann et al. (eds.),
The Impact of EU Accession on the Legal Orders of New Member States and (Pre-)Candidate Countries
(T.M.C. Asser Press 2006) p. 303, 304 et seq.
dictatorship169 – was expressed most explicitly by the second Lisbon judgment of

the Czech Constitutional Court. The Czech judges articulately rejected the peti-
tioners’ demand to denominate an abstract catalogue of non-transferrable rights
deduced from the Czech eternity clause under Article 9.2 in connection with
Article 1.1 of the Czech constitution:170
[T]he petitioners ask the Constitutional Court to set ‘substantive limits to the trans-
fer of powers’, and (...) attempt to formulate these themselves, evidently inspired by
the decision of the German Constitutional Court dated 30 June 2009 (...) which provides
such a catalogue in point 252 (...).
However, the Constitutional Court does not consider it possible, in view of the position
that it holds in the constitutional system of the Czech Republic, to create such a catalogue
of non-transferrable powers and authoritatively determine ‘substantive limits to the
transfer of powers’, as the petitioners request. It points out that it already stated [refer-
ence to Lisbon I] that ‘These limits should be left primarily to the legislature to specify,
because this is a priori a political question, which provides the legislature wide discretion’
[reference to Lisbon I, para 109]. Responsibility for these political decisions cannot be
transferred to the Constitutional Court; it can review them only at the point when they
have actually been made on the political level.
For the same reasons, the Constitutional Court does not feel authorised to formulate
in advance, in an abstract context, what is the precise content of Article 1.1 of the
Constitution, as requested by the petitioners, supported by the president, who wel-
comes the attempt “in a final list to define the elements of the ‘material core’ of the
constitutional order, or more precisely, of a sovereign democratic state governed by
the rule of law” (...).
(...) This does not involve arbitrariness, but, on the contrary, restraint and judicial
minimalism, which is perceived as a means of limiting the judicial power in favour of
political processes, and which outweighs the requirement of absolute legal certainty
(...). The attempt to define the term ‘sovereign, unitary and democratic state governed
by the rule of law, founded on respect for the rights and freedoms of the man and
of citizens’ once and for all (as the petitioners, supported by the president, request)
would, in contrast, be seen as an expression of judicial activism, which is, incidentally,
consistently criticized by certain other political figures.171
The reasoning of the Czech court raises questions that go beyond the sole aspect
of constitutional limits to European integration as it addresses the more funda-
¹69 For the historic background of the German eternity clause, see M. Herdegen, ‘Article 79’, in
T. Maunz and G. Dürig (eds.), Grundgesetz – Kommentar, 59th edn. (loose-leaf ) (Beck 2010) paras.
63 et seq.
¹70 According to Art. 9.2 of the Czech constitution, the ‘substantive requisites of the demo-
cratic, law-abiding State may not be amended.’ According to Art. 1.1, the ‘Czech Republic is a
sovereign, unitary and democratic, law-abiding State, based on respect for the rights and freedoms
of man and citizen.’
¹7¹ Czech CC Treaty of Lisbon II, supra n. 22 at paras. 110-113 (emphasis added).
mental question regarding the relationship between law and the political. To frame
it differently, at the heart of its reasoning, the Czech court raises the question of
institutional choice in the sense of to what extent a court should (re-)write the
constitution by means of interpretation and where it should leave the decision to
the legislator. The answer given is clearly in favour of the political process. Does
the second Lisbon judgment of the Czech court bear the hand-writing of those
authors who had underlined the limits of judicial reasoning on earlier occasions,
particularly within the context of the European Arrest Warrant?172
Anyway, as regards the determination of constitutional limits to further steps
of European integration, the comparative analysis reveals that the Conseil consti-
tutionnel and the Czech Constitutional Court throw the ball back into the po-
litical arena.173 In contrast, the Bundesverfassungsgericht claims to be competent to
determine limits to European integration which can only be overcome by super-
seding the German Basic Law. While the Conseil constitutionnel and the Czech
Constitutional Court stay within the existing constitutional system by referring
to the legislator or the constituent authority, the Bundesverfassungsgericht tends to
transcend the constitutional order (by which it is itself constituted) when it refers
to a pre-constitutional right to give oneself a constitution.174
Challenges to the applicability of EU law

Alongside the manifestation of substantial constitutional limits to future develop-
ments of EU (primary) law, the Lisbon decisions also address challenges to the
applicability of EU law in force. Some of the courts particularly highlight their
review power in cases of alleged ultra vires acts and violations of the so-called
national constitutional identity.
Ultra vires review

As far as national constitutional courts claim to be competent for ultra vires review
– that is to review if an act of EU law has transgressed the Union’s competences
and to declare it inapplicable in case it does175 – the courts agree in principle that
restrictive conditions must apply to its exercise.
¹7² See in particular J. Komárek, supra n. 11 at p. 38 et seq. within the context of the European
Arrest Warrant cases.
¹7³ This does not mean, however, that the Czech CC would pass the buck to the legislator in all
cases. But it holds that the ‘interference by the Constitutional Court should come into considera-
tion as ultima ratio’ only, see Czech CC Treaty of Lisbon I, supra n. 19 at para. 109.
¹74 See again German BVerfG Treaty of Lisbon, supra n. 21, para. 179.
¹75 ‘Ultra vires act’ is here understood in the narrower sense of the word, in contrast to the
wider concept which covers also infringements of other principles of EU law than only the principle
In the Lisbon judgment, the Bundesverfassungsgericht refined its previously de-

veloped176 concept of ultra vires review on several crucial points.177 The court
underlined that the decision to declare an act of EU law inapplicable within the
framework of an ultra vires review is reserved to the constitutional court alone.178
In a surprising obiter dictum, the court even proposed to the legislator to introduce
a new proceeding before the court ‘tailor made’ for this kind of review.179 But the
most important novelty was substantive: that the court limited ultra vires review
to ‘obvious transgressions’ and to cases where ‘legal protection cannot be obtained’
at EU level.180
In its Honeywell decision of 6 July 2010, the Bundesverfassungsgericht confirmed
this restrictive approach, albeit at the cost of having a dissenting opinion. The case
dealt essentially with the question of whether the German Federal Labour Court
could lawfully align its jurisprudence with the Mangold judgment of the Court of
Justice.181 According to the Bundesverfassungsgericht, ultra vires review requires a
prior reference to the Court of Justice in terms of procedure.182 In terms of substance,
an act of Union law may only be declared inapplicable if the breach of compe-
tences is ‘sufficiently qualified’. The fulfilment of this criterion depends on a
double test which will produce affirmative results only in very exceptional, not to
say hypothetical circumstances. Firstly, the act in question must constitute an
evident violation of competences. Secondly, the impugned act must entail a sig-
nificant impact on the system of distribution of competences between the member
states and the Union to the detriment of the former. In brief, the violation must
be obvious and simultaneously change structurally the division of competences
between the Union and the member states.183
Concerning possible ultra vires acts committed by the judicial branch and the
Court of Justice in particular, the Bundesverfassungsgericht affirmed in Honeywell
the ‘respect for the Union’s own methods of justice’ and even highlighted that the
of conferral, such as the fundamental rights for example. For the different notions, see F.C. Mayer,
Kompetenzüberschreitung und Letztentscheidung (C.H. Beck 2000) p. 24 et seq.
¹76 German BVerfG Maastricht, supra n. 74 at p. 187 et seq. The claim to exercise an ultra vires
review could already be identified previously in Case 2 BvR 255/69 Lütticke, order of 9 June 1971,
BVerfGE 31, 145, 174 and Case 2 BvR 687/85 Kloppenburg, order of 8 April 1987, BVerfGE 75,
223, 235.
¹77 This cannot be dealt with in detail here.
¹78 Ibid., para. 241.
¹79 At the moment there is no political majority which would opt in favour of such a solution,
which might even, for good reasons, give impetus to an infringement procedure under Art. 258
TFEU (ex-Art. 226 TEC).
¹8¹ ECJ, Case C-144/04 Mangold, judgment of 22 Nov. 2005.
¹8² German BVerfG Honeywell, supra n. 16 at para. 60.
¹8³ Ibid., Para 61.
Court of Justice had a ‘right to tolerance of error.’184 The question of whether the
Court of Justice transgressed the Union’s competences in its famous Mangold
judgment, was expressly left open by the German court,185 as it took the view that,
in any case, the second requirement of the double test was not fulfilled: ‘neither a
new field of competences was created for the Union to the detriment of the Mem-
ber States, nor was an existing competence expanded with the weight of a new
establishment.’
This restrictive approach should, however, not overshadow the fact that the Hon-
eywell case marks the first time that a national constitutional court actually under-
took an ultra vires review,186 even if the German court rejected the constitutional
complaint as unfounded.
The ultra vires claim is a prominent example for cross-border migration187 of
constitutional ideas. Originally developed by the Bundesverfassungsgericht, it was
taken up by the Danish Highest Court in its Maastricht decision,188 the Polish
Constitutional Tribunal in its decision on the Accession Treaty189 and also the
Czech Constitutional Court in its Lisbon decisions.190 The Polish and the Czech
decisions even drew explicitly on the German Maastricht decision for inspiration.191
Although not being equally sophisticated in terms of dogmatics, the jurispru-
dence of the Danish and the Czech courts principally correspond to the restrictive
course of the Bundesverfassungsgericht. Both courts underline that ultra vires review
must be exercised only in exceptional circumstances.192 A similar conclusion can
¹84 German BVerfG Honeywell, supra n. 16 at para. 66.

¹85 The question was whether the ECJ could derive a ‘general principle of the prohibition of
discrimination based on age’ from the constitutional traditions common to the Member States and
from their international agreements.
¹86 One could have interpreted as early as the Kloppenburg decision – supra n. 176 – in that way.
¹87 For the concept cf. S. Choudhry, ‘Migration as a new metaphor in comparative constitu-
tional law’, in id. (ed.), The Migration of Constitutional Ideas (CUP 2006) p. 1 et seq.
¹88 Danish HC, Case I 361/1997 Carlsen v. Rasmussen, judgment of 6 April 1998.
¹89 Polish CT Accession Treaty, supra n. 9 at points 10.3 and 4.5.
¹90 Czech CC Treaty of Lisbon I, supra n. 19 at paras. 120, 139 and 216; Treaty of Lisbon II, supra
n. 22 at para. 150 referring to the first decision.
¹9¹ This expression is used by the Czech CC in its first Lisbon decision, supra n. 19 at para. 111.
The Czech CC regularly refers to other constitutional courts. One of the most striking examples in
this respect is its second sugar quota case of 2006 in which it quotes the assembled prominence of
European landmark decisions relating to the principle of primacy, see Czech CC, Case Pl ÚS 50/04
sugar quotas II, decision of 8 March 2006, point VI.A.
¹9² Cf. supra n. 188 and n. 190.
be deduced from the jurisprudence of the Polish Constitutional Tribunal,193 even

despite some rhetorical differences.194
In brief, the Lisbon jurisprudence and subsequent cases confirm that the con-
stitutional courts which claim to be competent for ultra vires review construe it as
an exceptional review of last resort.
Identity review
The picture is more heterogeneous in relation to claims to review whether an act
of EU law violates the national constitutional identity. This begins with the term
‘identity’. Whereas the concept of ‘ultra vires act’ can be defined in a relatively
clear manner, the notion of identity remains obscure.195 But although (or just
because?) it runs risk of drifting away in the cloudy spheres of nebulosity, it nev-
ertheless seems to have the potential of becoming a universal term196 of European
constitutional law.
The notion of ‘constitutional identity’ is used by several national supreme ju-
risdictions as a synonym for constitutional core principles protected against the
primacy of EU law.197 While this alone is hardly revolutionary,198 the concept of
constitutional identity has recently turned out to be a genuine phenomenon of
multilevel-constitutionalism. Some of the courts no longer rely exclusively on
national constitutional law, but also on the new framed identity-clause in EU law
(Article 4.2 TEU) in order to justify the protection of the said core principles.
The first decision interesting in this respect is the declaration of the Spanish
Constitutional Tribunal on the Constitutional Treaty. For the Spanish Tribunal,
¹9³ In that sense also S. Biernat, supra n. 9 at para. 46.

¹94 The Czech CC distances itself expressly from the Polish CT: ‘The Polish Constitutional Tri-
bunal, for example, expressly rules out the jurisdiction of the Court of Justice to evaluate the limits
of conferral of competences on the EU, as, according to the Tribunal, that is a question of interpre-
tation of domestic constitutional law. Although, in terms of the dogmatics of domestic constitu-
tional law, we can agree with that conclusion to a certain extent, it is questionable whether it is
necessary to formulate it as sharply as the Tribunal did’ (Czech CC Treaty of Lisbon, supra n. 19 at
para. 139).
¹95 See J.-H. Reestman, supra n. 30 at p. 374 et seq. and A. v. Bogdandy, ‘Europäische und
nationale Identität: Integration durch Verfassungsrecht?’, 62 VVDStRL (2003) p. 156 and 164.
¹96 For the concept of universal terms in a judicial context, see M. Poiares Maduro, ‘Contrapunc-
tual Law’m in N. Walker (ed.), Sovereignty in Transition (Hart 2003) p. 501, 527 et seq.
¹97 See also F.C. Mayer, supra n. 31, 36.
¹98 For the use of the term ‘identity’ within the context of constitutional core principles, see
already German BVerfG, Case BvL 52/71 Solange I, order of 29 May 1974, BVerfGE 37, 271, 279
et seq. and Case 2 BvR 197/83 Solange II, order of 22 Oct. 1986, BVerfGE 73, 339, 375 f with
reference to the jurisprudence of the Italian CC (in particular Italian CC, Case 183/73 Frontini,
decision of 18 Dec. 1973) which does not use the term, but follows a comparable approach in
substance.
the identity clause (formerly planned as Article I-5 of the Constitutional Treaty)
is one of the key arguments to assume that an act of EU law which violates the
fundamental principles of the Spanish constitution would automatically also
constitute an infringement of European Union law and would therefore be sanc-
tioned already by Union law itself.199 To frame it differently, the Spanish Tribunal
sees the identity-clause as one of several ‘inbuilt breaks’ of the Treaties which ensure
that the Tribunal will never get into the position to have to declare an act of EU
law inapplicable in Spain, because before it reaches the Tribunal the infringement
will already have been sanctioned on the Union level.200 This is why the Spanish
Constitutional Tribunal can refer its own review power into the realm of the hy-
pothetical.
The identity-clause is also highlighted by the French Conseil constitutionnel in
its decision on the Constitutional Treaty. The Conseil concluded from a combined
reading of the (formerly planned) primacy clause and the identity clause that the
principle of primacy is, on the one hand, acknowledged, but, on the other hand,
also potentially limited. According to the Conseil, the Constitutional Treaty, and
particularly the close proximity of Articles 1-5 and 1-6 thereof, show that it in no
way modifies the nature of the European Union, nor the scope of the principle of
the primacy of Union law as duly acknowledged by Article 88-1 of the Constitution,
and confirmed by the Constitutional Council in its decisions.201
Among the jurisprudence referred to, a decision stands out in which the Conseil
constitutionnel stated that the obligation to transpose a directive follows not only
from EU law but also from French constitutional law (Article 88-1) and non-
transposition would only be possible on the grounds of an expressly contrary
provision of the French constitution.202 This constitutional entrenchment of the
obligation to transpose a directive allows the Conseil to sanction the enforcement
of EU law by means of French constitutional law and simultaneously to limit this
obligation by colliding constitutional provisions. In a decision of 2006 the Conseil
constitutionnel specified these potential limits as those regarding French ‘constitu-
tional identity’. According to the Conseil constitutionnel, the ‘the transposition of
¹99 Spanish CT Constitutional Treaty, supra n. 8, point II-3. In this sense, see the case note of
F. Castillo de la Torre, 42 CMLR (2005) p. 1169, 1195 et seq. and A.C. Becker, ‘Vorrang versus
Vorherrschaft’, EuR (2005) p. 353, 355.
²00 Cf. F. Castillo de la Torre, 42 CMLR (2005) p. 1169, 1193 and 1201.
²0¹ French CC Constitutional Treaty, supra n. 12 at para. 13.
²0² French CC E-Commerce, supra n. 13 at para 7. See also Case No. 2004-498 DC Bioethics,
decision of 29 July 2004, para 4. This reasoning has been aptly described as a legal osmosis, cf.
F. Chaltiel, ‘Constitution française, constitution européenne, vers l’osmose des ordres juridiques?’,
488 RMC (2005) p. 280 et seq.
a Directive cannot run counter to a rule or principle inherent to the constitu-

tional identity of France, except when the constituting power consents hereto.’203
While it is not clear what exactly falls under the ‘constitutional identity’ of France
– one could argue that this concept covers only those principles specific to the
French constitutional order, such as the principle of secularism –204it is definite
that the constitutional identity in the jurisprudence of the Conseil constitutionnel
does not constitute an inalienable limit to European integration. Instead, the final
decision whether or not infringement of French constitutional identity should be
allowed rests with the constituting power.
There are several differences between the approach of the French Conseil con-
stitutionnel and the German Bundesverfassungsgericht as regards the concept of
constitutional identity.205 But the most important is the following. While in Ger-
many the Bundesverfassungsgericht grants the constitutional identity absolute
protection under the eternity clause, in France the constituent authority maintains
the right to adapt the French constitution. If the French Conseil constitutionnel
ever held that the implementation of a directive would infringe the constitu-
tional identity of France, then the constitution could be revised.
In Germany, on the contrary, the notion of constitutional identity is bound to
Article 79.3 of the Basic Law, which shields the substantial core of the constitution
even against a revision. On this basis the Bundesverfassungsgericht claims to be
competent to review whether an act of EU law infringes Germany’s constitu-
tional identity and is thus inapplicable within Germany. The court holds that
otherwise
the fundamental political and constitutional structures of sovereign Member States,

which are recognised by Article 4.2 first sentence Lisbon TEU, cannot be safe-
guarded in any other way. In this respect, the guarantee of national constitutional
identity under constitutional and under Union law go hand in hand in the European
legal area. The identity review makes it possible to examine whether due to the action
of European institutions, the principles ... declared inviolable in Article 79.3 of the
Basic Law, have been violated. This ensures that the primacy of application of Union
law only applies by virtue and in the context of the constitutional empowerment
that continues in effect.206
²0³ French CC Information Society, supra n. 13 at para. 19.

²04 Cf. in that sense the conclusions of the former rapporteur public at the French Conseil
d’Etat M Guyomar in case Arcelor, RTDE (2007) p. 378, 385.
²05 This cannot be dealt with in detail here. For details, see the brilliant analysis of J.-H. Reest-
man, supra n. 29 at p. 384 et seq., particularly p. 388-390.
²06 German BVerfG Treaty of Lisbon, supra n. 21 at para. 240 (emphasis added).
The picture of ‘hand in hand’ is deceiving. It alludes to mutual loyalty, which

would indeed make sense as the new identity clause is placed directly alongside
the principle of mutual loyalty enshrined in Article 4.3 TEU. However, as to the
legal consequences of the Bundesverfassungsgericht’s approach, the metaphor of
‘pulling and tearing’ would be more adequate. The German court construes the
identity clause as a mere reflex or fingerprint of (national) sovereign statehood. In
other words, Article 4.2 TEU is seen as a declarative affirmation of the Bundesver-
fassungsgericht’s alleged right to decide unilaterally on the inapplicability of EU
law within Germany. Once the constitutional identity is touched, the game is over.
Even if the German court underlines that the exercise of this review power must
be reconciled with the constitutional principle of ‘friendliness towards European
Law’207 and seems to indicate in its Honeywell decision that the restrictive proce-
dural requirements set up for the ultra vires review might also apply for the iden-
tity review,208 one must strongly doubt that the Bundesverfassungsgericht’s approach
corresponds to an adequate understanding of Article 4.2 TEU.
First, according to the Bundesverfassungsgericht’s conception, everything depends
on the content or non-content of the term identity. Is it now possible to circum-
vent the Solange II decision and the Banana decision209 by claiming that the es-
sential core of a fundamental right – protected as part of the constitutional
identity – has been violated?210 If so, the reference to the identity-clause might
constitute opening Pandora’s box.211
Second, even if the German Lisbon judgment does not re-open the Solange
saga,212 its approach remains, again, blind to the multi-levelled structure of Euro-
pean constitutionalism. If it is true that the identity-clause aims at the protection
of specificities and core principles of national constitutions, then EU law cannot
determine what ‘constitutional identity’ is. To that extent Article 4.2 TEU is open
to varying evaluations of national authorities, including national courts. But the
²07 Ibid.
²08 German BVerfG Honeywell, supra n. 16 at para. 59: ‘According to the legal system of the
Federal Republic of Germany, the primacy of application of Union law is to be recognised and it is
to be guaranteed that the control powers which are constitutionally reserved for the Federal Con-
stitutional Court are only exercised in a manner that is reserved and open towards European law.’
²09 German BVerfG, Case 2 BvL 1/97 Banana-Market, order of 7 June 2000, BVerfGE 102,
147 et seq.
²¹0 In Case 1 BvR 256/08 et al. Data retention, judgment of 2 March 2010, para. 218 the Ger-
man BVerfG stated: ‘It is part of the constitutional identity of the Federal Republic of Germany
that the citizens’ enjoyment of freedom may not be totally recorded and registered, and the Federal
Republic must endeavour to preserve this in European and international connections.’ The BVerfG
here referred explicitly to the identity control under para. 240 of the Lisbon judgment. In its Lisbon
judgment the BVerfG also identified the principle of nulla poena sine culpa as being an integral part
of the German constitutional identity, German BVerfG Treaty of Lisbon, supra n. 21 at para. 364.
²¹¹ See Mayer, supra n. 31, 39 et seq. concluding that in some way any problem of protection of
fundamental rights could thus be treated as a problem of constitutional identity too.
²¹² In this sense Thym, supra n. 38 at p. 1807.
question of content must be separated from the question of normative relevance.

Even if Article 4.2 TEU is an auto-limitative response of EU law to the claims of
national jurisdictions that certain core principles of national constitutions are not
subject to the principle of primacy, then the question of how far EU law limits its
own claim of primacy still remains a question of EU law. This is a logical conse-
quence of the formal separation of national and supranational law. Consequently
the extent to which the identity claim is normatively relevant within the realm of
EU law is a question to be decided by EU law and thus by the Court of Justice.
In other words, according to our interpretation, while national constitutional
courts may decide on the content of constitutional identity, the Court of Justice
decides if and to what extent the identity claim prevails over (hypothetically)
conflicting principles of EU law.
Seen in this way, Article 4.2 appears as an integration clause on Union-level
ensuring the legal permeability of EU law with regard to national constitutional
law.213 By means of the identity clause, EU law revokes to some extent – and not
unlimited – its own claim of primacy within its scope of application. Hence, the
task of protecting national constitutional identity is, under EU law, not a task
reserved for national courts. It is instead distributed between supranational and
national level.
In this respect it is important to note that Article 4.2 contains, according to its
wording, an obligation for the EU to ‘respect’ the member states’ national iden-
tity, not to generally outweigh other principles of EU law against it. Otherwise,
we would be back to Pandora’s box.214 The deliberations of the European Conven-
tion confirm the view that the identity clause is not framed as a unilateral ‘deroga-
tion clause’ on which the Court of Justice215 not be competent to rule.216 The role
of the Court of Justice as a ‘first line of defence’ seems to be presumed also by the
Spanish Constitutional Tribunal in its declaration on the Constitutional Treaty as
far as it assumes that a violation of the core principles of the Spanish constitution
would not occur, because such a violation would simultaneously constitute an
infringement of EU law sanctioned by the Court of Justice.217
²¹³ In legal terms, permeability can be defined as the capacity of a given legal order to limit its
own claim of normative exclusivity in order to enable legal rules or principles which emanate from
a formally separated legal order to integrate. See in detail Wendel, supra n. 3, chs. 1 and 13.
²¹4 See also A.G. M. Poiares Maduro, Case C-213/07 Michaniki, opinion of 28 Oct. 2008,
para. 32 et seq.
²¹5 The ECJ was, under the former Art. 46 of the TEU-Nice, not competent to interpret the
identity-clause under the former Art. 6.3 TEU-Nice.
²¹6 CONV 375/1/02 REV 1, p. 11: ‘In the latter respect it was noted that the provision was not
a derogation clause. The Member States will remain under a duty to respect the provisions of the
Treaties. ... Were the Court of Justice to be given power with respect to such article in a future
“basic treaty of constitutional significance”, the Court could be the ultimate interpreter of the pro-
vision if the political institutions went beyond a reasonable margin of appreciation.’
²¹7 See also F. Castillo de la Torre, 42 CMLR (2005) p. 1169, 1201.
However, once the protection of constitutional identity has been identified as

a genuine task of multi-level constitutionalism in which the preliminary reference
procedure assures the necessary dialogue between the Court of Justice and the
national courts, the question arises of whether the Court of Justice has taken up
its mission so far. Is the Court of Justice really capable of a multi-level conception
of the identity clause? This premise would at least entail the Court of Justice han-
dling the identity clause as a justiciable provision of EU law, now that the Court
of Justice is legally competent to interpret it.218 However, even if the Advocates-
General have brought up the question of constitutional identity regularly before
the Court of Justice in recent times,219 the Court of Justice remains silent about
the possible impact of Article 4.2 TEU and thus risks of handing over the reins to
the national courts.220
Conclusion: A new quality of comparative dialectics

To conclude, the Lisbon decisions reveal significant differences not only in terms
of procedure, but first and foremost in terms of substance. While the German
Bundesverfassungsgericht’s Maastricht decision became an often quoted leading case
in cross-border perspective, it is unlikely that its Lisbon judgment will have a
similar effect.
The comparative analysis reveals instead that the Bundesverfassungsgericht is the
only constitutional court in Europe which – based on a conception of democracy
bound existentially to statehood – demands a prior constitutive assent of na-
tional parliament in all possible cases of so-called dynamic treaty provisions. The
Bundesverfassungsgericht is also the sole court to declare itself competent to spell
out a constitutional eternity clause in a detailed (albeit apodictic) manner, enlist-
ing whole areas within which the future conferral of competencies to the EU could
bear a high risk of violating national constitutional identity.
But beyond the mere fact of apparent differences, the Lisbon judgments reveal
a new quality of judicial dialogue and comparative exchange of national courts.
Indeed, it is not a new phenomenon that national constitutional courts stay in
close, albeit informal, contact221 and refer to each other sporadically in their de-
cisions.222 But in the Lisbon saga, the frequency of such cross-references is rather
²¹8 Cf. supra n. 215.

²¹9 See particularly the opinion of A.G. M. Poiares Maduro in the case Michaniki, supra n. 213
at para. 32. For more examples, see Wendel, supra n. 3, ch. 13 point II-2-a.
²²0 An example how the protection of constitutional identity might work in practice is the
Omega-case, ECJ case C-36/02 Omega, judgment of 14 Oct. 2004. However, the term ‘identity’ is
not used within this decision.
²²¹ Cf. for example the Conference of European Constitutional Courts, <www.lrkt.lt/confer
ence5.html>.
²²² Cf. supra n. 198.
high. References can be found chronologically in the decisions of the Czech Con-
stitutional Court, the German Bundesverfassungsgericht, the Hungarian Constitu-
tional Court and the Polish Constitutional Tribunal. Most of these references are
punctual affirmations used to support a specific argument as far as it is congruent
with the approach taken by another court. These references do not imply that the
judgments referred to serve as a normative basis, but rather reflect a comparative
‘inspiration’ as the Czech Constitutional Court has called it. The possibility of
being inspired in this way is increased by the fact that the supreme jurisdictions
show more and more tendency to publish their ‘European’ judgments also in
English translations.223
However, particularly the Czech Constitutional Court has shown that judicial
dialogue on horizontal level can involve critical comparative reasoning as well. Its
second Lisbon decision is a clear signal that the interaction and mutual influence
of national jurisprudence does not necessarily lead to the reception of a specific
judgment in the jurisprudence of other courts, but may in fact evoke express rejec-
tion. By objecting openly to some of the key arguments of its German counterpart,
the Czech Court in its second Lisbon judgment disproved the commonly expressed
idea that constitutional courts of Eastern and Central European countries stick to
the Bundesverfassungsgericht as a sort of archetype of constitutional court.
Does this contravene the idea of a common European constitutional law? No,
on the contrary. With its second Lisbon decision, the Czech court in fact demon-
strated the dialectical prospects of comparative reasoning within the multi-levelled
cooperation of European constitutional courts.224 Its assessment of the German
Lisbon decision should be understood as a constructive impulse that may even
foster the debate. The Czech Constitutional Court examined the German decision
intensely and came to the conclusion that some of the central arguments and as-
sumptions of this decision – particularly those regarding the principle of sover-
eignty, the European Parliament and the judicial handling of the constitutional
eternity clause – were not sufficiently convincing or transferable to be taken up
within the framework of the Czech constitutional order. In doing so, the Czech
Constitutional Court itself made an important contribution to the discussion
about the scope and content of a common European constitutional law. Com-
parative dialectics between national constitutional courts in Europe have reached
a new level of quality.

²²³ The German BVerfG issued an English translation for the first time in the European Arrest
Warrant case. In the Lisbon- and Honeywell decisions a translation was already available the day the
decision was published.
²²4 See A. Voßkuhle, supra n. 35.

Lisbon Before The Courts Comparative Perspectives

Uploaded by

Copyright:

Available Formats

Lisbon Before The Courts Comparative Perspectives

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Lisbon Before The Courts Comparative Perspectives

Uploaded by

Copyright:

Available Formats

96 Mattias Wendel EuConst 7 (2011)

Lisbon Before the Courts:

European Constitutional Law Review, 7: 96–137, 2011

¹ I. Pernice, ‘The Treaty of Lisbon: Multilevel constitutionalism in Action’, 15 CJEL (2009)

primacy (primacía) and supremacy (supremacía).8 Similarly, the decision of the

¹7 French CC, Case 2007-560 DC Treaty of Lisbon, decision of 20 Dec. 2007.

Supreme Court.26 Alongside the decisions of national supreme jurisdictions, there

A preliminary word about comparison

Procedural background and general overview

Preventive treaty reviews: France and the Czech Republic

France – necessity of a prior constitutional amendment

Czech Republic – two decisions, no amendments

Ex post treaty reviews: Hungary and Poland

Hungary – ‘the use of the ex ante review would be desirable’

59 Ibid., para. 116.

Poland – protest during the oral hearing

65 Ibid., point IV.2.

Individual complaints and petitions: Germany, Austria and Latvia

Germany – the right to vote as a catalyst to a full-scale review

a right to democratic self-determination, to free and equal participation in the state

8¹ Ibid., paras. 168, 181.

Austria – total inadmissibility

Latvia – a middle way

9³ Austrian CC Treaty of Lisbon I, supra n. 18, point II.2.

Differences of institutional self-conception?

99 Ibid., points 9 and 13.

constitutionality. The Latvian Constitutional Court in turn went a middle way as

Legitimizing the EU: A case of multi-level democracy?

Parliamentary assent to the future application of ‘dynamic treaty provisions’

¹04 For judicial ‘acts of self-authorisation’, see – in a historical-comparative perspective –

Five categories of so-called105 ‘dynamic treaty provisions’ can be distinguished.

Similarities and peculiarities

¹05 Cf. German BVerfG Treaty of Lisbon, supra n. 21 at para. 239.

National v. multi-levelled democracy

The TL would introduce additional democratic guarantees, namely, national Parlia-

The Czech Constitutional Court additionally draws on the jurisprudence of the

If one was asked to trace back the arguments of the Bundesverfassungsgericht to a

Who demands prior parliamentary assent: division of powers I

Popular vote as a requirement for ratification?

Constitutional limits and judicial reservations

Substantive constitutional limits to future developments of EU law

Two categories of constitutional limits

Particularly sensitive for the ability of a constitutional state to democratically shape

¹5² Hungarian CC Treaty of Lisbon, supra n. 23, point IV.2.3.2.

Who determines constitutional limits: division of powers II

¹59 German BVerfG Treaty of Lisbon, supra n. 21 at para. 259.

dictatorship169 – was expressed most explicitly by the second Lisbon judgment of

Challenges to the applicability of EU law

Ultra vires review

In the Lisbon judgment, the Bundesverfassungsgericht refined its previously de-

¹84 German BVerfG Honeywell, supra n. 16 at para. 66.

be deduced from the jurisprudence of the Polish Constitutional Tribunal,193 even

¹9³ In that sense also S. Biernat, supra n. 9 at para. 46.

a Directive cannot run counter to a rule or principle inherent to the constitu-

the fundamental political and constitutional structures of sovereign Member States,

²0³ French CC Information Society, supra n. 13 at para. 19.

The picture of ‘hand in hand’ is deceiving. It alludes to mutual loyalty, which

question of content must be separated from the question of normative relevance.

However, once the protection of constitutional identity has been identified as