Special Section: Opinion 2/13
“It’s the Autonomy, Stupid!” A Modest Defense of Opinion 2/13
on EU Accession to the ECHR, and the Way Forward
By Daniel Halberstam*
A. Introduction
The Court of Justice of the European Union has arrived! Gone are the days of hagiography,
when in the eyes of the academy and informed observers the Court could do no wrong.
The pendulum has finally swung the other way. The judicial darling, if there is one today, is
Strasbourg, not Luxembourg. Not hours had passed before the Court’s 258-paragraph long
Opinion 2/13
1
on the Draft Agreement on EU Accession to the European Convention on
Human Rights was condemned as “exceptionally poor.”
2
Critical voices have mounted
steadily ever since, leading to nothing short of widespread “outrage.”
3
*
Eric Stein Collegiate Professor of Law and Director, European Legal Studies, University of Michigan Law School.
Contact: dhalber@umich.edu.
1
Opinion Pursuant to Article 218(11) TFEU, C-2/13 (Dec. 18, 2014), http://curia.europa.eu/ [hereinafter Opinion
2/13].
2
See Steve Peers’s statement via Twitter dated December 18, 2014, available at
https://twitter.com/StevePeers/status/545523536551768064: “My summary of CJEU ruling on EU accession to
ECHR. Blog post coming later. Preview: an exceptionally poor judgment “
3
Walther Michl, Thou Shalt Have No Other Courts Before Me, VERFBLOG (Dec. 13, 2014),
http://www.verfassungsblog.de/en/thou-shalt-no-courts/. For a sampling of the numerous critical comments,
see, e.g., Leonard F.M. Besselink, Acceding to the ECHR Notwithstanding the Court of Justice Opinion 2/13,
VERFBLOG (Dec. 23, 2014) available at http://www.verfassungsblog.de/en/acceding-echr-notwithstanding-courtjustice-opinion-213/;
Pieter Jan Kuijper, Reaction to Leonard Besselink’s ACELG Blog, AMSTERDAM CENTRE FOR
EUROPEAN LAW AND GOVERNANCE (Jan. 6, 2014), available at http://acelg.blogactiv.eu/2015/01/06/reaction-toleonard-besselinks’s-acelg-blog/;
Tobias Lock, Oops! We did it again – the CJEU’s Opinion on EU Accession to the
ECHR, VERFBLOG (Dec. 18, 2014), available at http://www.verfassungsblog.de/en/oops-das-gutachten-des-eughzum-emrk-beitritt-der-eu/;
Aidan O’Neill, Opinion 2/13 on EU Accession to the ECHR: the CJEU as Humpty Dumpty,
EUTOPIALAW (Dec. 18, 2014), available at http://eutopialaw.com/2014/12/18/opinion-213-on-eu-accession-to-theechr-the-cjeu-as-humpty-dumpty/;
Steve Peers, The CJEU and the EU’s Accession to the ECHR: A Clear and Present
Danger to Human Rights Protection, EULAWANALYSIS, (Dec. 18, 2014), available at
http://eulawanalysis.blogspot.com.es/2014/12/the-cjeu-and-eus-accession-to-echr.html; S. Douglas-Scott,
Opinion 2/13 on EU accession to the ECHR: a Christmas Bombshell From the European Court of Justice, U.K. CONST.
L. BLOG (Dec. 24, 2014), available at http://ukconstitutionallaw.org; Mattias Wendel, Mehr Offenheit wagen! Eine
kritische Annäherung an das Gutachten des EuGH zum EMRK-Beitritt, VERFBLOG, (Dec. 21, 2014), available at
http://www.verfassungsblog.de/mehr-offenheit-wagen-eine-kritische-annaeherung-das-gutachten-des-eughzum-emrk-beitritt.
For the very few more positive voices, see, C. Barnard, Opinion 2/13 on EU Accession to the
ECHR: Looking for the Silver Linings, EULAWANALYSIS, (Dec. 16, 2015), available at
http://eulawanalysis.blogspot.com/2015/02/opinion-213-on-eu-accession-to-echr.html; A. Duff, The European
106 German Law Journal [Vol. 16 No. 01
The Court, for its part, has not made things easy. As with James Carville’s famous slogan to
which the title of this piece alludes,
4
an abrasive message may work better as an internal
catchphrase among friends than as an external message to an audience that needs
convincing. The tone of Opinion 2/13, and its many uncompromising interpretive moves
along the way, sure make it seem as though Strasbourg is not welcome in Luxembourg.
Wary of its younger and overburdened sibling, the CJEU seems intent on guarding its
privileged judicial position in Europe. Especially when compared to Advocate General
Kokott’s rather softer views ending in “Yes, but only if . . . ,” the Opinion’s harsh “No,
unless . . .” seems stark.
5
Opinion 2/13 appears all the more abrupt given the judges’ own
involvement in the process leading up to the proposed agreement.
6
I, too, disagree with a good deal in the Court’s Opinion. And as a critic on record of the
CJEU
7
as well as, in the pages of this Journal, the German Constitutional Court,
8
I have little
patience for judicial hagiography. But a whiff of quite the opposite is in the air. These days
it seems more tempting to follow the scent of weak prey, than to delve into the highly
intricate legal issues and resist moving in for the kill.
Upon careful and moderately charitable inspection, however, I argue that Opinion 2/13
warrants far more serious attention than its numerous critics suggest. As an initial matter,
Union is in Deep Trouble with its Top Court, BLOGACTIV.EU, (Jan. 7, 2015), available at
http://andrewduff.blogactiv.eu/2015/01/07/the-european-union-is-in-deep-trouble-with-its-top-court/; Jean
Paul Jaqué, Non à l’adhésion à la Convention Européenne des Droits de l’homme?,” DROIT UE, (Dec. 13, 2014),
available at: http://www.droit-union-europeenne.be/412337458/2394230/posting/texte; Henri Labayle, La
guerre des juges n’aura pas lieu. Tant mieux? Libres propos sur l’avis 2/13 de la Cour de justice relatif à l’adhésion
de l’Union à la CEDH, GDR-ELSJ.EU, (Dec. 22, 2014), available at: http://www.gdr-elsj.eu. Martin Sheinin, in turn,
steers clear of the substance, arguing in favor of three “mitigating circumstances” for the Court’s Opinion
“without any intention to defend the Opinion itself.” M. Sheinin, CJEU Opinion 2/13 – Three Mitigating
Circumstances, VERFBLOG, (Dec. 26, 2014), http://www.verfassungsblog.de/en/cjeu-opinion-213-three-mitigating-
circumstances.
4
James Carville, strategist to Bill Clinton in the 1992 election campaign, coined the campaign slogan: “It’s the
economy, stupid!” It started out even shorter: “The economy, stupid!” For obvious reasons, the phrase was
originally intended only as internal slogan, not for public consumption.
5
Compare View of Advocate General Kokott, Opinion Procedure 2/13, Case C-2/13 (June 13, 2014),
http://curia.europa.eu/ with Opinion 2/13.
6
CJEU Council Document 13714/10 (Sept. 17 2010).
7
Daniel Halberstam & Eric Stein, The United Nations, the European Union, and the King of Sweden: Economic
Sanctions and Individual Rights in a Plural World Order, 46 CMLR 13 (2009).
8
Daniel Halberstam & Christoph Möllers, The German Constitutional Court Says “Ja zu Deuschland”, 10 German
L.J. 1241 (2009).
2015] Opinion 2/13 – “It’s the Autonomy, Stupid!” 107
the opinion cannot simply be ignored.
9
It is binding on the Member States and the EU
institutions, none of which can proceed with accession without in good faith changing
either the Draft Agreement or the Treaty. It would also be a serious mistake for the
Member States to follow the related suggestion to sign a protocol declaring that accession
shall take place “notwithstanding Article 6(2) Treaty on European Union, Protocol No 8
relating to Article 6(2) of the Treaty on European Union and Opinion 2/13 of the Court of
Justice of 18 December 2014.”
10
Adopting such a protocol as part of the accession agreement would mean circumventing
Article 48 of the Treaty on European Union (“TEU”), which lays down the exclusive way to
amend the EU’s treaties. At least since Kadi,
11
this is no longer a constitutional option. A
protocol adopted as part of an international treaty would not qualify as a treaty revision to
override the Court’s opinion within the meaning of Article 218(11) of the Treaty on the
Functioning of the European Union (“TFEU”). The adoption of any “Notwithstanding
Protocol” would have to follow the full procedures of Article 48 TEU, including—unless the
European Parliament agrees otherwise—a full Convention.
But even if properly crafted and properly adopted, a blanket dismissal of the Court’s
concerns would still be throwing out the baby with the bathwater. The bracing exchange of
pluralism, of which I am a strong proponent, lacks value and values if constitutionalism is
not part of the mix. The internal constitutional perspective of actors considering external
legal claims does not undermine pluralism. To the contrary, constitutionalism provides a
framework for legitimacy of the exercise of public power. As a result, constitutionalism
supplies the terms on which the pluralist contestation takes place. As I have argued
repeatedly elsewhere, constitutionalism supplies the “grammar of legitimacy” that
governs the pluralist contest by insisting that power always vindicate a combination of
“‘voice,’ ‘rights,’ and ‘expertise.’”
12
We must, therefore, never forget the role that
constitutionalism plays in a pluralist constellation.
But the current critics did just that: They rushed to embrace Strasbourg while forgetting
about the constitutional dimension of EU governance along the way. This singular focus on
international human rights regimes can be misleading. Participation in international human
9
Pieter Jan Kuijper, Reaction to Leonard Besselink’s ACELG Blog, AMSTERDAM CENTRE FOR EUROPEAN LAW AND
GOVERNANCE, (Jan. 6, 2014), available at http://acelg.blogactiv.eu/2015/01/06/reaction-to-leonard-besselinks’s-
acelg-blog/.
10
Besselink, supra note 3. In a carefully and interesting editorial forthcoming in the Common Market Law Review
(privately shared with the author), Leonard Besselink has since clarified his position, discussing the
notwithstanding protocol in terms of a proper amendment under TEU art. 48.
11
Kadi and Al Barakaat Int’l Fndtn v Council and Comm’n, Case C‐402/05 P and C‐415/05 P, 2008 ECR 461.
12
D. Halberstam, Local, Global, and Plural Constitutionalism: Europe Meets the World, in THE WORLDS OF EUROPEAN
CONSTITUTIONALISM 150, 170-171 (G. de Búrca & J. Weiler eds., Cambridge University Press, 2012).
108 German Law Journal [Vol. 16 No. 01
rights regimes should be encouraged and, indeed, brought about. But signing on to a
particular rights regime ought not to come at the expense of the constitutional nature of
the EU’s legal order, which is geared to vindicating all three constitutional values, including
rights. The EU may therefore sign on to the ECHR as an extension, but not as a substitution,
of its own project of constitutional governance.
In sum, the EU’s constitutional engagement with the world leaves ample space for hard
pluralist contestation. But we must first understand the constitutional element of the EU’s
side of the contest. It is in this plural constitutionalist spirit that I reconstruct the Court’s
objections to the Draft Agreement.
B. Advisory Opinions
We will get into the nitty-gritty of EU law very quickly. A comparativist at heart, however, I
cannot help but begin by noting that this kind of opinion would have been impossible in
the United States. U.S. federal courts, including the United States Supreme Court, do not
opine on matters of constitutional law outside the context of what Americans call a “case
or controversy.”
13
When Secretary of State Thomas Jefferson asked the United States
Supreme Court for constitutional advice on a delicate matter of treaty interpretation in the
spring of 1793, the Justices brushed him off. This was not a judicial matter, they said, as it
did not involve two adverse parties locked in a real case. As such, the Supreme Court did
not have jurisdiction to opine on the matter. The Chief Justice wrote back to President
Washington: “We exceedingly regret every event that may cause embarrassment to your
administration, but we derive consolation from the reflection that your judgment will
discern what is right.”
14
The European Union Treaties clearly take the opposite approach, allowing advisory
opinions under Article 218(11) TFEU well in advance of any real case. And it seems that
here, it’s the opinion given, not the opinion refused, that is causing some rather serious
embarrassment in Brussels. Opinion 2/13 is binding, which means either the Draft
Agreement must be changed or the EU Treaties must be amended. The CJEU quite likely
regrets the embarrassment Opinion 2/13 has caused. Indeed, in the light of the Court’s
answers, some members of the Commission might wish they’d never asked the Court in
the first place. After all, no Treaty provision required them to do so.
13
U.S. Const. Art. III, Sec. 2.
14
Letter from John Jay to George Washington, in 3 THE CORRESPONDENCE AND PUBLIC PAPERS OF JOHN JAY 488–89
(Johnston ed., 1891).
2015] Opinion 2/13 – “It’s the Autonomy, Stupid!” 109
C. How Did We Get Here?
Brussels seems shocked by Opinion 2/13. Some complain that when acting as observer in
the drafting of the agreement,
15
the Court failed to indicate the concerns that now trip up
the result of that process in the courtroom. Others lament that Luxembourg did not leak
any part of its decision to those in power before dropping the Christmas bombshell.
Putting aside whether the Court is estopped from complaining or should warn actors about
impending rulings,
16
were there really no warning signs?
Just take Opinion 2/94,
17
which was the last time the Court was asked to opine on whether
EU accession to the European Convention on Human Rights (“ECHR”) was legal. EU
accession had been bandied about since the late 1970s, but shortly after the Maastricht
Treaty officially included respect for fundamental rights in the Treaty on European Union,
things got more serious.
The Council wanted to know whether EU accession was compatible with the Treaties.
Opinion 2/94 cleverly deflected that question, choosing to declare instead that the EU did
not have the competence to join the ECHR. The Court opined that joining the ECHR would
exceed the general rounding out competence of what was then Article 235 of the Treaty
on European Community. And why was that? Opinion 2/94 points right back to what are
best understood as serious concerns about the compatibility of accession:
34. Accession to the Convention would . . . entail a
substantial change in the present Community System
for the protection of human rights in that it would
entail the entry of the Community into a distinct
international institutional system as well as integration
of all the provisions of the Convention into the
Community legal order.
35. Such a modification of the system for the
protection of human rights in the Community, with
equally fundamental institutional implications for the
Community and for the Member States, would be of
15
See supra note 6.
16
On the second complaint, recall that Justice Felix Frankfurter reportedly leaked the U.S. Supreme Court’s
impending Ex Parte Endo decision to the Roosevelt administration which, in turn, led the President to announce
the closing of the U.S. detention camps before the ruling declared them unlawful. This allowed the Administration
to save face. It is also likely why so few remember the Supreme Court’s decision in Ex Parte Endo today. See
Patrick O. Gurthridge, Remember Endo?, 116 HARV. L. REV. 1933 (2003).
17
Case C-2/94 (Mar. 28, 1996), http://curia.europa.eu/.
110 German Law Journal [Vol. 16 No. 01
constitutional significance and would therefore be such
as to go beyond the scope of Article 235. It could be
brought about only by Treaty Amendment.
18
What did the political branches do in response? Two things. First, they added a bare
mandate to Article 6 TEU that “[t]he Union shall accede to the European Convention for
the Protection of Human Rights and Fundamental Freedoms.”
Second, they added a series of what you might call stand-still provisions. One was in Article
6 TEU: “Such accession shall not affect the Union’s competences as defined in the
Treaties.” The others were placed in Protocol 8 of the TEU and TFEU; demanding that
accession shall (a) “preserv[e] the specific characteristics of the Union and Union law,” (b)
“not affect the competences of the Union or the powers of its institutions,” (c) not “affec[t]
the situation of Member States in relation to the European Convention,” and (d) not
“affect Article 344 of the Treaty on the Functioning of the European Union” (which
provides that the Member States will resolve their Treaty-related disputes exclusively by
the mechanisms provided for in the Treaties).
19
Put another way, after Opinion 2/94 said that accession to the ECHR would entail changes
of constitutional proportions in the European Union’s legal order, the political actors
turned around with a mandate to join the ECHR, demanding at the same time that there be
no significant changes in the European Union’s legal order. The amendment also made it a
requirement that nothing change in the Member States’ relationship to the ECHR.
Threading the needle of accession on these terms would be a very difficult task. To be sure,
the Court might well have read the accession mandate somewhat more broadly. And as I
argue below, there is indeed room for political actors to push back against the strictest
reading of Opinion 2/13. But putting all this and the soundness of Opinion 2/94 itself aside,
it cannot come as a complete shock to serious observers of the Court that, this time
around, the CJEU balked, once again, at the question of compatibility.
There were other warnings, too. Add to Opinion 2/13 the Kadi decision,
20
and the Court’s
Opinion 1/09 on the Unified Patent Court.
21
Kadi confirmed the Court’s view that EU law is
18
Id. at paras. 34–35.
19
Consolidated Version of the Treaty on the Functioning of the European Union art. 15, May 9, 2008, 2008 O.J. (C
115) 47 [hereinafter TFEU].
20
Kadi, Case C‐402/05 P and C‐415/05 P.
21
CJEU Case C-1/09 (Mar. 8, 2011), http://curia.europa.eu/ [hereinafter Opinion 1/09].
2015] Opinion 2/13 – “It’s the Autonomy, Stupid!” 111
autonomous not only vis-à-vis Member State law, but also vis-à-vis international law.
22
As a
consequence, no international legal obligation can alter core principles of EU constitutional
law. The Patent Court Opinion, in turn, held that the application of EU law must remain in
the hands of the EU judiciary, which includes national courts but excludes courts in which
Member States and non-Member States participate together.
23
Both were strong
assertions of the autonomy of EU law and the necessity of maintaining the integrity of the
EU’s constitutional architecture.
No warning? I think not.
24
Rightly or wrongly, there were strong signals that the Court
would not bend existing principles to accommodate accession. Nothing, in the Court’s
view—not even a Treaty Amendment that broadly commands accession to the ECHR—may
compromise the autonomy of the European Union legal order.
D. The Court’s Opinion in Brief
The Court’s Opinion 2/13 holds the Draft Agreement to be incompatible with the Treaties
for five basic reasons, some with multiple subparts. The following is a quick and
straightforward summary.
First, the Draft Agreement disregards the specific characteristics of EU Law in several ways.
Article 53 ECHR reserves to the Contracting Parties the power to lay down higher
standards, whether in their national laws or in international agreements. The CJEU,
however, has already interpreted the similarly worded Article 53 of the EU Charter of
Fundamental Rights (“CFR”) as precluding higher Member State standards that undermine
the primacy, unity, and effectiveness of EU Law.
25
Opinion 2/13 now similarly objects to
Article 53 ECHR, noting that this provision should be coordinated with Article 53 CFR, as
interpreted by the Court of Justice, again to preserve the primacy, unity, and effectiveness
of EU law.
26
The Draft Agreement further runs afoul of the specific characteristics of EU law in that the
ECHR obligation of Member States to check each other in their observation of fundamental
rights is incompatible with the mutual trust that Member States must show one another,
22
Cf., e.g., Gráinne de Búrca, The European Court of Justice and the International Legal Order after Kadi, 51 Harv.
Int. L.J. 1 (2010); Halberstam and Stein, supra note 7.
23
Opinion 1/09, at paras. 78, 80 and 89. Cf., e.g., Roberto Baratta, National Courts as ‘Guardians’ and ‘Ordinary
Courts’ of EU Law: Opinion 1/09 of the ECJ, 38:4 LEG. ISSUES OF ECON. INTEGRATION 297 (2011).
24
I am unlikely alone in this assessment. For a good discussion prior to Opinion 2/13, see Gráinne de Búrca, The
Road Not Taken, 105 AJIL 649 (2011).
25
Judgment in Melloni, Case C-399/11 (Apr. 5, 2013), http://curia.europa.eu/.
26
See Opinion 2/13 at paras. 187–190.
112 German Law Journal [Vol. 16 No. 01
especially in the area of Freedom, Security, and Justice. Even if there are extraordinary
circumstances in which such trust might no longer be applicable, the Court says EU law
controls those situations, and may do so to the exclusion of any other law.
27
Protocol 16 of the ECHR—to which the EU would not become a party—would also violate
the specific characteristics of EU law. The Protocol allows Member State high courts to ask
Strasbourg for an advisory opinion on a question of ECHR interpretation. After accession,
this would potentially circumvent the EU’s own preliminary reference system.
28
Second, the Court holds that the Draft Agreement would violate Article 344 TFEU—which
grants exclusive jurisdiction over disputes regarding EU law to the method of settlement
provided for in EU law—by allowing Member States or the EU to initiate proceedings
against one another under Article 33 ECHR.
29
Third, the co-respondent mechanism runs afoul of EU Treaty requirements in several ways.
Allowing the European Court of Human Rights (“ECtHR”) to review the plausibility of a
request from the Member States or the EU to become a co-respondent would improperly
allow the ECtHR to decide on a question of EU law.
30
Allowing the ECtHR to hold corespondents
jointly liable may run contrary to the Member State’s reservation regarding
that particular obligation.
31
And it is also unacceptable to allow the ECtHR to decide, on the
basis of the reasons given by the parties, to place responsibility on only one of the corespondents,
as this would infringe on the CJEU’s exclusive jurisdiction to rule on the
division of responsibilities among the EU and the Member States.
32
Fourth, the prior involvement procedure fails to guarantee that the Court of Justice can
decide all questions of EU law. The Court holds that the Draft Agreement must be changed
to inform the EU, and to allow the appropriate EU institution to trigger the prior
involvement of the CJEU in cases where it is doubtful whether the CJEU has already
decided the question of EU law at issue in the case.
33
Furthermore, the prior involvement
procedure must be available for the CJEU to examine not only the validity of provisions of
27
See id. at paras. 191–195.
28
See id. at paras. 196–199.
29
See id. at paras. 201–214.
30
See id. at paras. 222–225.
31
See id. at paras. 226–228.
32
See id. at paras. 229–234.
33
See id. at paras. 238–241.
2015] Opinion 2/13 – “It’s the Autonomy, Stupid!” 113
secondary EU Law and the interpretation of primary law, but also regarding the
interpretation of provisions of secondary law.
34
Finally, the Court declares the Draft Agreement incompatible with the specific
characteristics of EU law insofar as it would entrust judicial review, including of
fundamental rights and ECHR compliance, of some Common Foreign and Security Policy
(“CFSP”) matters exclusively to the ECtHR.
35
E. Evaluating the Court’s Objections: A Constitutional Approach
Let’s get one thing out of the way. The President of the Court of Justice has been quoted as
saying at the FIDE Conference in 2014: “The Court is not a human rights court. It is the
Supreme Court of the European Union.” Commentators quote this passage with the
suggestion that it demonstrates the Court’s failure to take human rights seriously.
36
The
argument echoes a rather old debate,
37
recently renewed by suspicions about the Court’s
bona fides regarding labor rights after Viking and Laval
38
and asylum rights after N.S. and
Abdullahi.
39
Rights lapses at the Court and anywhere else must be condemned, but there is
nevertheless a good deal of respectable truth to the President’s statement. After
significant prodding from Member State high courts, the Court has come a long way from
its early days of dismissing such claims. Today, the Court sees fundamental rights as at the
very heart of EU law, indeed as a precondition of the legality and legitimacy of the entire
legal order.
40
The Union is broadly committed to protecting rights, even to joining
international human rights agreements—as well it should. Nonetheless, solicitude for
international human rights agreements comes with a caveat. The Court will show solicitude
for international human rights agreements only insofar as these agreements and their
34
See id. at paras. 242–247.
35
See id. at paras. 249–257.
36
E.g., S. Douglas-Scott, Opinion 2/13 on EU Accession to the ECHR: A Christmas Bombshell from the European
Court of Justice, U.K. CONST. L. BLOG, (Dec. 24 2014), available at http://ukconstitutionallaw.org.
37
See J.H.H. Weiler & N.J.S. Lockhart, ‘Taking Rights Seriously’ Seriously: The European Court of Justice and its
Fundamental Rights Jurisprudence, 32 C.M.L. Rev. Parts I & II (1995); J. Coppel & A. O’Neill, The European Court of
Justice: Taking Rights Seriously?, 29 C.M.L. Rev. 669 (1992).
38
See, e.g., VIKING, LAVAL, AND BEYOND (M. Freedland & J. Prassl eds., 2014).
39
See, e.g., Steve Peers, Tarakhel v. Switzerland: Another Nail in the Coffin of the Dublin System?, EU Law Analysis
Blog, (Nov. 5, 2014), available at http://eulawanalysis.blogspot.com. Cf. infra notes 79–101 and accompanying
text.
40
See, e.g., Kadi, Case C‐402/05 P and C‐415/05 P at paras. 283, 284.
114 German Law Journal [Vol. 16 No. 01
various institutional arrangements do not undermine the constitutional architecture of the
European Union.
As the theory of constitutional pluralism holds, the constitutional architecture of the
European Union need not displace all other claims of constitutional authority from within
or even beyond the EU.
41
As far as the Member States are concerned, the Court has long
participated in the give and take of constitutional pluralism, i.e., the multiplicity of claims
to constitutional authority among the Member States and the European Union and their
respective high courts. As far as potential claims of constitutional pluralism from actors
beyond the EU are concerned, by contrast, the Court has still been rather more reluctant
to play the same game.
42
But even if we push for a pluralist vision of the relationship
among the CJEU, Member State high courts, and the ECtHR in Strasbourg,
43
we must
nonetheless first understand the constitutional element of the EU’s side of the bargain.
If constitutionalism is the overarching dimension along which the preservation of the
specific characteristics of the European legal order must be judged, another involves the fword.
Put aside for the moment the endless disputes about defining federalism.
44
The web
of mutual obligations among the European Union and its Member States goes well beyond
that seen in any other international organization to date. Nonetheless, the EU and its
Member States do not necessarily replicate the specific relationship between the central
government and the component states of a traditional federal state. The same goes for the
relationship among the Member States themselves, as compared to the signatory parties
to any other international organization, on the one hand, and to the component states of a
traditional federal state, on the other. To understand the Court’s concerns, we will
therefore carefully draw on the theory and practice of federalism without ignoring the
special characteristics of the European Union.
Taken together, constitutionalism and federalism leave international law and the European
Union in a quandary as they approach each other in Strasbourg. There is near universal
agreement that the EU is not a state. But this negative conclusion does not tell us how,
positively and systematically, to conceive of the relationship between the European Union
and the Member States, and among the Member States themselves, especially as this
constitutional bundle of joined legal systems approaches international law.
41
See, e.g., CONSTITUTIONAL PLURALISM IN THE EUROPEAN UNION AND BEYOND (Mateij Avbelj & Jan Komarek, eds., 2012)
(with pluralist contributions from Halberstam, Komarek, Kumm, Maduro, and Walker).
42
Halberstam and Stein, supra note 7; De Burca, supra note 7.
43
Andreas Vosskuhle, Der europäische Verfassungsgerichtsverbund, 106 Staatlicheit im Wandel, (Dec. 12, 2009),
http://www.sfb597.uni-bremen.de/pages/pubApBeschreibung.php?SPRACHE=de&ID=146.
44
S. Rufus Davis, THE FEDERAL PRINCIPLE: A JOURNEY THROUGH TIME IN QUEST OF A MEANING (1978).
2015] Opinion 2/13 – “It’s the Autonomy, Stupid!” 115
We know from Kadi that the Court considers the European Union to be a constitutional
entity. But this still leaves open how the EU’s peculiar federal-type nature should generally
affect its relationship with international law. This overarching question is related to, but
different from, the very old but ever fresh subject of competences and mixed
agreements.
45
It goes one step deeper. How does the deep federal-type structure of the
EU’s constitutional order affect even those agreements to which the EU is a full-fledged
independent party? The terms of this aspect of the engagement are still being set.
Reading the Court’s opinion in light of this overarching question, we see that the Court
properly identifies some problems while suggesting sensible remedies. On other matters,
the Court properly identifies certain problems, but nonetheless seems misguided on the
remedy. Even on this approach, there is at least one element of the Opinion that seems
misguided on the underlying substantive complaint. Finally, we shall see along the way
that the Court’s concerns can all be addressed while preserving space for a pluralist give
and take between Luxembourg and Strasbourg.
I. The Co-Respondent Mechanism and the Prior Involvement of the Court
The Court’s concerns about the co-respondent mechanism and about the prior
involvement procedure contained in the Court’s third and fourth main points above are
both well founded. And both demand a remedy along the lines the Court suggests. These
issues need not be discussed at great length here given that the Advocate General’s
opinion—which has been as universally cheered as the Court’s has been condemned—fully
agrees with the Court. The only difference between the Advocate General and the Court
on these two points is rhetorical: the Advocate General concludes with “yes, but only if”
whereas the Court says “no, unless”. The different reception of the two opinions on these
issues shows that a difference in tone goes a long way.
To state matters briefly, the Advocate General agrees that becoming a co-respondent
cannot depend on a plausibility check on the part of the ECtHR.
46
The agreement must
clarify that joint responsibility will not affect any reservations made by the Contracting
Parties,
47
and that allocating responsibility to one or the other party, even according to the
reasons given by the parties, is beyond the scope of the ECtHR’s jurisdiction.
48
She
concludes that the Draft Agreement and the accompanying explanations must be changed
to preserve the autonomy of EU law.
45
Cf., e.g., MIXED AGREEMENTS REVISITED: THE EU AND ITS MEMBER STATES IN THE WORLD (Christoph Hillion & Panos
Koutrakos, eds., 2010).
46
View of Advocate General Kokott, supra note 5 at para. 229–235.
47
Id. at para. 260–265.
48
Id. at para. 175–179.
116 German Law Journal [Vol. 16 No. 01
The Advocate General further agrees that the EU must be reliably informed so as to trigger
a co-respondent application,
49
and that the prior involvement of the Court must also be
engaged where there is any doubt on whether the CJEU has resolved the EU law question
at issue in the case.
50
Finally, the AG agrees that the Draft Agreement must be clarified to
establish that the subject matter of such prior involvement cannot be limited to the
determination of compatibility, but must also encompass the interpretation of secondary
law.
51
Were the ECtHR to decide—however minimally—on whether to allow the EU or a Member
State to become a co-respondent, Strasbourg would directly or indirectly be deciding on
the distribution of competences among the EU and the Member States. From the
perspective of international law, the question of state responsibility is surely considered to
be distinct from the internal law of the state or international organization to be judged.
52
So Strasbourg can, indeed, lay claim to authority to decide which of two Contracting States
bears responsibility for any given violation of the ECHR. But from the perspective of EU
law, determining the responsibility for a violation and the competence to provide a remedy
are nonetheless both strictly questions of EU law.
53
The clash between these two perspectives is the stuff of pluralism, which need not (and
perhaps should not) be destroyed. But there is a trick to this. Call it the art of ambiguity:
pluralism could have been preserved by failing to specify who ultimately decides questions
of responsibility. We know this all too well from our experience with constitutional
pluralism elsewhere. By eliding the question of ultimate authority, the question of “who
decides” can become a matter of mutual accommodation. We see this in the EU’s
“internal” constitutional pluralism, i.e., the EU’s relationship with its Member States.
54
In
the same way, determining the plausibility of whether the EU should be a co-respondent
(as well as allocating final responsibility) could have been left to the mutual
accommodation between Strasbourg and Luxembourg.
49
Id. at para. 222–228.
50
Id. at para. 180–184.
51
Id. at para. 130–135
52
See, e.g., Draft Articles on Responsibility of States for Internationally Wrongful Acts, with Commentaries, [2001]
2 Y.B. Int’l L. Comm’n 2, ¶ 77, U.N. Doc. A/CN.4/SER.A/2001/Add.1 (Part 2); Draft Articles on the Responsibility of
International Organizations, with Commentaries, Rep. of the Int’l Law Comm’n, 63rd Sess., Apr. 26–June 3, July 4–
Aug. 12, 2011, ¶ 88, UN Doc. A/66/10.
53
E.g., View of Advocate General Kokott, supra note 5 at para. 232.
54
I borrow the term “internal” constitutional pluralism from Miguel Maduro. See M. Poiares Maduro, Interpreting
European Law: Judicial Adjudication in the Context of Constitutional Pluralism, 2 EUR. J. LEGAL STUD. 1 (2007).
2015] Opinion 2/13 – “It’s the Autonomy, Stupid!” 117
Unfortunately, however, the Draft Agreement (plus explanations) chose a different path. In
the matter of who decides on responsibility in the context of the joint respondent
mechanism, the draft treaty sought, rather unhelpfully, to clears things up by brokering a
compromise. It gave the EU the greater part of the power to determine responsibility,
while allowing Strasbourg to check only the plausibility of the EU’s claim and to allocate
responsibility on the basis of the arguments put forward by the EU and whatever Member
State found itself before the ECtHR.
The Draft Agreement’s compromise is what caused the problem. From the perspective of
EU law, the EU cannot sign a document that formally grants the ECtHR the power to decide
these questions of EU law (even if only at the margins.) Signing such a document would be
signing away the CJEU’s power to determine what the law of the Union is. As a
constitutional matter, this is not possible. Moreover, after accession, the ECHR would
become binding EU law, and the ECtHR’s decisions interpreting the ECHR (including joint
responsibility questions, if that were allowed) would take on great significance within the
EU’s legal order.
The two other objections need no further elaboration. It seems evident that the CJEU must
retain authority to interpret, not just to invalidate, EU law, and that it must retain this
power in all cases in which there is any doubt on whether the CJEU has had the
opportunity to do so. Let’s please just not make the same mistake when fixing the Draft
Agreement on this score. Do not specify that the ECtHR may decide whether such doubt
exists.
55
That may create the same problem as assigning jurisdiction over the plausibility
check to the ECtHR.
II. Article 344 and the Exclusivity of EU Adjudication of Member State Disputes
This issue, the Court’s second main concern, is easy on substance, but far less so on
remedy. Here, too, the Advocate General agrees with the Court on the substantive
concern. Proceedings under Article 33 ECHR by one Member State against another, or
between the EU and any one Member State, would violate the exclusivity provision of
Article 344 TFEU.
56
The AG and the Court differ somewhat, however, on the question of remedy. The AG
submits that an express statement in the Draft Agreement is unnecessary. As the AG
55
The revised accession agreement would best not specify who determines whether such “doubt” exists. Once
the agreement specifies the institution that has ultimate authority over whether “doubt” exists, the Court will,
once again, insist that this power be given to the CJEU.
56
TFEU art. 344 provides: “Member States undertake not to submit a dispute concerning the interpretation or
application of the Treaties to any method of settlement other than those provided for therein.”
118 German Law Journal [Vol. 16 No. 01
properly notes, some international agreements do contain an express statement regarding
the priority of regional dispute settlement systems,
57
but such priority statements are not
common in international law agreements (including several the EU has signed in the past).
If the Court is really worried, the AG suggests it could make accession subject to a binding
unilateral declaration on the part of the EU and the Member States that neither the EU nor
the Member States would invoke the Article 33 ECHR interstate dispute provision when
the subject matter of the dispute falls within the scope of EU law.
58
Here, too, the difference between the celebrated views of the AG and the condemned
Opinion of the Court is rather slim. The AG says no action is needed but does suggest a
unilateral binding declaration on the part of the Member States rejecting ECtHR
jurisdiction. The CJEU, by comparison, says “only the express exclusion of the ECtHR’s
jurisdiction under Article 33 of the ECHR over disputes between Member States and the EU
in relation to the application of the ECHR within the scope ratione materiae of EU law
would be compatible with Article 344 TFEU.”
59
To be sure, perhaps the most faithful
reading of this passage in the Court’s opinion is that a unilateral declaration does not
suffice. And yet, a unilateral declaration submitted in the form of a reservation and
accepted by the ECtHR might just count as such an express exclusion. Whether the Court
would ultimately accept such a unilateral binding agreement may depend on how many of
the Court’s other complaints are properly heard by the political actors.
To understand what remedy might be needed, we must better understand why the Court
thought there was a problem in the first place. After all, several commentators plausibly
point out that the Article 33 ECHR issue looks much more like an EU problem than an ECHR
problem. If Member States violate EU law by resorting to the ECHR’s dispute resolution
mechanism, why turn to the ECHR to save the Union from the illegal behavior of its own
actors? Also, one might ask, even if this problem were real, does the problem not already
exist in the absence of accession?
This is a powerful critique, but let me nonetheless try to rehabilitate the Court just a bit. As
several commentators have pointed out, the Court remarks grandly that the Article 33
ECHR dispute resolution mechanism is problematic because “the very existence of such a
possibility undermines the requirement set out in Article 344 TFEU.”
60
But commentators
have uniformly failed to notice the very next paragraph of Opinion 2/13, which adds: “This
is particularly so since, if the EU or Member States did in fact have to bring a dispute
57
See United Nations Convention on the Law of the Sea, 1833 U.N.T.S. 397, art. 282 (Dec. 10, 1982).
58
View of Advocate General Kokott, supra note 5 at para. 120.
59
Opinion 2/13 at para. 213.
60
Id. at para. 208.
2015] Opinion 2/13 – “It’s the Autonomy, Stupid!” 119
between them before the ECtHR, the latter would, pursuant to Article 33 of the ECHR, find
itself seised of such a dispute.”
61
What does this mean?
If one Member State were to take another Member State, or the EU, before the court in
Strasbourg in violation of Article 344 TFEU,
62
the Strasbourg court’s jurisdiction would not
be optional. On its face, the ECHR would demand that the ECtHR entertain the suit. To be
sure, one might plausibly argue that a Member State was thereby acting in violation of
good faith or in abus de droit.
63
The AG puts the point this way: “An objection of
inadmissibility could . . . be raised before the ECtHR in respect of any inter-State case that
was nevertheless initiated.”
64
But whether general principles of law or principles of general
international law would allow the Strasbourg court to reject the dispute because of a
violation of a treaty other than the ECHR is not clear. If the EU were to violate its own
constitutional treaty by bringing a suit in the ECtHR that does not belong there, it would
seem rather simple to argue that the suit does not belong there. But if a Member State,
which is party to the TEU and TFEU on the one hand, and to the ECHR on the other hand,
brings a suit in one forum in violation of the other forum’s exclusivity rules, the case for
dismissal will likely be a much closer call.
Even if Member States currently can misuse Strasbourg in just this way, accession would
worsen the problem in two ways. First, the EU currently cannot be party to such abusive
suits, whether in bringing suits or being sued, simply because the ECtHR lacks jurisdiction
over the EU. After accession, however, an abusive suit could also be brought by and
against the EU itself. Second, by acceding to the ECHR in a way that potentially allows for
such suits, the EU is becoming party to the agreement that serves as the basis for such
abusive suits.
The fact that other, existing international treaties the EU has signed, such as the WTO, for
instance, may potentially create similar problems is somewhat beside the point. For
example, a suit by one Member State against another or against the EU within the scope of
EU law may well be far less sustainable as a good faith use of the WTO’s dispute resolution
mechanism than a suit among Member States or by a Member State against the EU under
the ECHR. But even if agreements the EU has already signed create this problem as well,
that should not estop the CJEU from objecting to this one. After all, nobody has argued
61
Id. at para. 209.
62
This is no far-fetched hypothetical, as the Mox Plant litigation illustrates. See Judgment in Commission v. Ireland
(‘Mox Plant’), CJEU Case C-459/03, EU:C:2006:345 (May 30, 2006).
63
See, e.g., Appellate Body Report, United States—Import Prohibition of Certain Shrimp and Shrimp Products, ¶
158, WT/DS58/AB/R (Oct. 12, 1998) (adopted Nov. 6, 1998).
64
View of Advocate General Kokott, supra note 5 at para. 115.
120 German Law Journal [Vol. 16 No. 01
that the CJEU has ever considered this problem in the past, or specifically approved of this
aspect of a prior treaty.
In summary, if the ECtHR were indeed required to hear the dispute under the
circumstances just described, then the CJEU’s concern may seem rather more plausible. It
is one thing not to require Strasbourg to help keep rogue EU Member States in check. It is
quite another to ask the Luxembourg to approve of an agreement extending the reach of a
provision that, under certain factual circumstances, mandates a violation of EU law.
This reconstruction, in my view, renders the CJEU’s substantive concern far more plausible
than might appear at first blush. I might nonetheless have encouraged the Court to show
somewhat greater tolerance and to entertain this risk. But from the internal, constitutional
perspective of the CJEU, these are not implausible concerns.
So what about the remedy?
A binding declaration of the Member States along the lines suggested by the AG would
seem sufficient to solve this problem. Such a binding declaration would almost certainly
allow the ECtHR to dismiss any action brought under Article 33 ECHR in contravention of
Article 344 TFEU as a violation of good faith. With the existence of such a declaration, the
Member States would be acting against the intent they themselves publicly expressed as
part of the EU’s accession. This places the commitment not to sue one another in
Strasbourg in violation of Article 344 TFEU into a domain clearly cognizable by the ECtHR.
Again, such a declaration is probably not what the Court had in mind, but the Court might
be convinced to accept it depending on how the Court’s other concerns are addressed.
Notice that this binding declaration vindicates the Court’s constitutional premise, but
ultimately preserves an important element of pluralism in the arrangement with
Strasbourg. Such a declaration would allow the dismissal of routinely abusive suits and
thereby address the Court’s legitimate concerns. At the same time, however, in an extreme
case, a litigant might argue that the EU’s system of remedies had failed completely, and
that the situation was therefore no longer covered by the declaration. The binding
declaration would thus eliminate the core problem while still leaving room for a pluralist
system of global checks and balances.
III. Protocol 16 on the Optional Advisory Opinion Procedure
Let us take the Advocate General as our point of departure again. This time, though, the
Advocate General differs from the Court both in the perception of the problem and in the
proposed solution. As we shall see, the Court may have the better of the argument on
substance, even if not on remedy.
2015] Opinion 2/13 – “It’s the Autonomy, Stupid!” 121
On the substance, the Advocate General says the threat of Protocol 16 to the preliminary
reference procedure would exist with or without EU accession to the ECHR. The AG, along
with some commentators, suggests that accession cannot be unlawful as a result of
Protocol 16. After all, so this claim goes, even absent accession, Member States may turn
to the ECtHR with questions about fundamental rights, notably with questions about ECHR
provisions that could have been framed as questions about the EU Charter of Fundamental
Rights. The problem, then, is not one created by the Draft Agreement.
65
This rejection of concerns about Protocol 16 is, once again, too quick. Keeping the
constitutional autonomy of the EU’s legal order in mind, the problem of Protocol 16 is a
real problem, and a distinct problem after accession. Here’s why.
Before accession, an ECHR question is mostly that: a question about the interpretation of
the ECHR. Article 52(3) FRC refers to the ECHR for content, but that provision does not
incorporate the ECHR into EU law as a legally operative norm. After accession, by contrast,
an ECHR question indeed becomes a question of EU law, at least insofar as the question
implicates the EU’s own ECHR obligations. This means that the advisory opinion process
creates a real risk of undermining the EU’s preliminary reference procedure after
accession, and far more so than before.
We must, once again, move away from an exclusively human rights focused interpretive
approach and towards constitutional analysis in order to see the significance of the
problem more clearly. Without EU accession, if a Member State court asks the ECtHR
instead of the ECJ, it might well be asking the wrong question. That is, the Member State
court should have asked Luxembourg about the Charter instead of asking Strasbourg about
the Convention. From the perspective of the CJEU, this is bad, of course. But at least the
Member State court is not thereby mixing up the two courts and the legal systems over
which the two courts have jurisdiction.
After accession, by contrast, the mistaken Member State court would be asking Strasbourg
about the Convention, when it should be asking Luxembourg about the Convention. Recall
that accession turns the Convention into an integral part of EU law, binding on all actors
and institutions of the Union.
66
This means that after accession, the CJEU must be
considered the authoritative interpreter of the ECHR-as-EU-law for all matters that fall
within the scope of EU law. After accession, therefore, the mistaken Member State court
that rings up Strasbourg instead of Luxembourg would be asking a non-EU court a question
of EU law.
65
View of Advocate General Kokott, supra note 5 at para. 140.
66
Haegeman v. Belgian State, Case 181/73, EU:C:1974:41, para. 5.
122 German Law Journal [Vol. 16 No. 01
To be sure, the mistaken Member State could be thought of as asking Strasbourg the
Convention question purely as a matter of Convention law (conventionality), not as a
matter of EU law (legality). But it is not clear this distinction matters much. First, after
accession the Member States should be taking their lead from the CJEU not the ECtHR
even on matters of conventionality, as long as they fall within the scope of EU law. Second,
accession may turn the distinction between conventionality and legality into an almost
metaphysical one insofar as the EU’s status as a Contracting Party bears internally on
judging the legality of EU (secondary) law. Because the Convention will have become
internally binding EU law—even if the advisory opinion itself would apply a margin of
discretion and not be binding
67
—the difference between legality and conventionality (in
judging secondary EU law) may become rather slight in the eyes of Member State courts
adjudicating fundamental/human rights challenges.
For related reasons, accession may spur Member State courts to make more such
“mistaken” requests for advisory opinions from Strasbourg than before. Once the ECHR
binds the EU, a Member State court might see the advisory opinion as a shield against
(secondary) EU law the Member State court does not like. Prior to accession, the Member
State court would have to invoke that Member State’s full sovereignty and independent
obligations under the Convention to resist the commands of an “offending” EU law. After
accession, by contrast, the Member State court might be encouraged just to play off its
own less-favored EU obligation (a secondary provision of EU law the Member Court
dislikes) against the higher ranking EU obligation (compliance with the ECHR) after a
conversation with Strasbourg. The fact that Strasbourg might involve Luxembourg in
answering this question via the prior involvement procedure adds insult to injury from the
CJEU’s perspective because the entire proceeding should be in Luxembourg, subject to the
jurisdiction of the CJEU.
Once again, considering the question from a constitutional perspective makes the Court’s
concerns quite a bit more plausible. But, do we need to change the agreements as a
result?
The AG says that any threat is alleviated by the obligation of Member State courts of last
resort under Article 267 TFEU to refer open questions of EU law to Luxembourg.
68
Given
the pervasive possibility of Member State court “confusion” spelled out above, however,
this may not suffice. In addition, it may be useful to distinguish between the mere
possibility of a violation of EU law and signing an agreement that requires or virtually
ensures the ECtHR’s participation in the violation of EU law. Here, the apparent obligation
on the part of the Strasbourg court to hear such advisory opinions is indeed less
67
See Protocol No. 16 to the European Convention for the Protection of Human Rights and Fundamental
Freedoms, Feb. 2013, C.E.T.S. No. 214, art. 5.
68
View of Advocate General Kokott, supra note 5 at para. 141.
2015] Opinion 2/13 – “It’s the Autonomy, Stupid!” 123
pronounced than in the case of Article 33 ECHR. Strasbourg can reject such requests but
must “give reasons” for doing so.
69
But nothing would currently tell Strasbourg to refuse a
request for an advisory opinion on an open ECHR question that falls within the scope of EU
law. A binding unilateral declaration would provide guidance and provide a solid legal
basis for the ECtHR to refuse jurisdiction when a Member State wrongly requests an
advisory opinion where the question falls within the competence of the CJEU.
As in the case of Article 344 TFEU, this solution of a binding unilateral declaration to the
Protocol 16 problem gives due consideration to the Court’s constitutional premise without
sacrificing pluralism along the way. A declaration would expressly commit the Member
States vis-à-vis the ECHR to using the EU’s procedures for questions concerning EU law. It
would thereby eliminate the potential legal requirement on the ECtHR to hear mistaken
questions that should be brought within the EU’s own legal system. And yet, unlike the
more explicit express exclusion possibly suggested by the Court, this would leave the door
open just a crack: In an extraordinary case where the EU’s own legal system had massively
failed in a structural sense, a Member State might argue before the ECtHR that there was
no EU legal process available, and hence the declaration inapplicable.
IV. The Puzzling Nature of both Article 53s and the Problem of “Higher Standards”
We are beginning to get into more difficult terrain. There has always been something
counterproductive and downright misleading about Article 53, whether in the ECHR or in
the Charter. Both provisions purport to ensure that nothing in the Charter or the
Convention, as the case may be, shall derogate from existing rights in the laws—or
constitutions—of the signatory or Member States and their international agreements.
These claims are deeply problematic.
On its face, the Article 53 claim that the Convention, or the Charter as the case may be,
does not override rights found elsewhere runs up against any minimally sophisticated
understanding of rights.
70
As soon as parties claim rights on both sides of an issue, no law
or legal institution can possibly protect one of these rights while completely deferring to a
69
Protocol No. 16 to the European Convention for the Protection of Human Rights and Fundamental Freedoms,
Feb. 2013, C.E.T.S. No. 214, art. 2.
70
Compare European Convention on Human Rights art. 53, opened for signature Nov. 4, 1950, C.E.T.S. No. 005
[hereinafter ECHR] (“Nothing in this Convention shall be construed as limiting or derogating from any of the
human rights and fundamental freedoms which may be ensured under the laws of any High Contracting Party or
under any other agreement to which it is a party.”) with Charter of Fundamental Rights of the European Union
art. 53, Dec. 18, 2000, 2000 O.J. (C 364) 1 [hereinafter CFR] (“Nothing in this Charter shall be interpreted as
restricting or adversely affecting human rights and fundamental freedoms as recognised, in their respective fields
of application, by Union law and international law and by international agreements to which the Union or all the
Member States are party, including the European Convention for the Protection of Human Rights and
Fundamental Freedoms, and by the Member States' constitutions.”).
124 German Law Journal [Vol. 16 No. 01
different law or legal institution on the definition and assessment of the other right
involved. Where rights claims are made on both sides, holding for one side inevitably
means rejecting, or balancing away, the other side’s claim. If a court sets limits on positive
gender-based action as a way of vindicating one person’s right to equal treatment, for
instance, it cannot remain infinitely open to a more broadly conceived right to genderbased
preferences. To take a concrete example from the case law, a Member State cannot
grant female job applicants a constitutional right to absolute and unconditional priority
over men whenever the woman has the same formal educational degree as the man
against whom she is competing. This kind of absolute and unconditional priority would
almost certainly violate what the CJEU has determined to be the man’s right to equality.
71
Similarly, a court cannot declare that you have, say, a right to free expression in a
particular case without also ruling on the right of competing claims to free expression or
dignity when such a competing right is in play.
72
So, for the Convention (or the Charter, as
case may be) to guarantee one right while claiming completely to preserve all others,
regardless of how they may be defined by the states, makes rather little sense. One might
have objected to Article 53—whether in the Charter or the Convention—on that ground
alone.
As far as Article 53 of the EU’s Charter of Fundamental Rights was concerned, it should
have been clear that the generic reservation of more expansive rights at the Member State
level was dead on arrival. Since the earliest days of the Community, and confirmed
repeatedly even after the Union expressed its concern for human rights, a Member State
was not allowed to object to EU law simply on the grounds that the EU measure violated
an idiosyncratic human right found in that Member State’s constitution.
73
The rights
reservation in Article 53 CFR could not possibly resurrect those claims. To be sure, there
can be—and there is—a certain give and take between the Member State high courts and
the CJEU on defining the extent of rights protection with the Union. Indeed, the EU has
long derived its own set of fundamental rights in one way or another from the Member
71
See, e.g., Eckhard Kalanke v Freie Hansestadt Bremen, Case C-450/93, [1995] ECR I-03051; Hellmut Marschall v
Land Nordrhein-Westfalen, Case C-409/95, [1997] ECR I-06363; Georg Badeck and Others, Case C-158/97, [2000]
ECR I-01875; Katarina Abrahamsson and Leif Anderson v Elisabet Fogelqvist, Case C-407/98, [2000] ECR I-05539;
H. Lommers v Minister van Landbouw, Natuurbeheer en Visserij, Case C-476/99, [2002] I-02891; Serge Briheche v
Ministre de l’Intérieur, Ministre de l’Éducation nationale and Ministre de la Justice, Case C-319/03, [2004]ECR I-
8807; Judgment in Commission of the European Communities v Hellenic Republic, Case C-559/07,
ECLI:EU:C:2009:198, 26 March 2009. 

72
For a German example, see the landmark judgment in Lüth, Bundesverfassungsgericht [BVerfG – Federal
Constitutional Court], Case No. 1BvR 400/51, 7 ENTSCHEIDUNGEN DES BUNDESVERFASSUNGSGERICHTS [BVerfGE] 198 (Jan.
15, 1958) (striking down defamation charge against applicant as violating his free speech rights). For a prominent
U.S. example, compare Catharine MacKinnon, Pornography as Defamation and Discrimination, 71 B.Y.U. Law Rev.
793 (1991) with American Booksellers Ass'n, Inc. v. Hudnut, 771 F.2d 323 (7th Cir. 1985), aff’d, 475 U.S. 1001
(1986) (striking down local anti-pornography statute as violating free speech).
73
See, e.g., Mannesmannröhren-Werke, Case T-112/98, [2001] E.C.R. II-729, para. 84.
2015] Opinion 2/13 – “It’s the Autonomy, Stupid!” 125
States’ constitutional and international law practice. But from the internal perspective of
the CJEU, the formal reservation of the unilateral power of the Member State to create any
additional fundamental/human rights they please was always ruled out. And so, sure
enough, the Court recently confirmed in Melloni that a Member State’s idiosyncratic
fundamental rights catalogue cannot undermine the primacy, unity, and effectiveness of
EU law.
74
Opinion 2/13 expresses the worry that the Member States might now use Article 53 of the
Convention to resurrect fundamental rights standards in defiance of Melloni. As a factual
matter, this is not an implausible concern. After all, Melloni, too, was a real case, not an
imagined one. There, too, one Member State sought to resist the application of EU law by
invoking an idiosyncratic constitutional right in conjunction with Article 53 FRC. Such a
case, then, might well arise under the Convention as well.
Regardless whether such a case might arise, however, the claim is legally unfounded for
one simple reason: the reservation in the Convention cannot create a power that did not
previously exist. Article 53 ECHR says that “[n]othing in this Convention shall be construed
as limiting or derogating from any of the . . . rights . . . which may be ensured under the
laws of any High Contracting Party or under any other agreement to which it is a party.”
75
This is not a power provision. It is a rule of construction that purports to limit the effect of
the remaining provisions of the Convention. It does not grant the Contracting Parties any
right they did not already have prior to signing on to the Convention. Therefore, Article 53
ECHR also cannot resurrect a power EU law has already denied the Member States. If
Member States today are denied the power to maintain “higher”
76
standards that violate
the primacy, unity, and effectiveness of EU law, then Article 53 of the Convention does not
and cannot give them that power in the future. If a Member State seeks to use the
Convention to impose a higher standard of rights, that “higher” standard of rights can and
will be reviewed under the Melloni doctrine just as it was before.
If the Court’s objection to Article 53 ECHR is unfounded, what about the Court’s proposed
remedy? Fortunately, the Court did not specifically demand one. The Court demands only
that “[i]n so far as Article 53 ECHR essentially reserves the power of the Contracting Parties
to lay down higher standards” that provision “should be coordinated with Article 53 of the
74
Melloni, Case C-399/11.
75
ECHR art. 53
76
I am generally puzzled by the phrase “higher” in this context, as it suggests some kind of linear metric of rights.
But that seems to be the way the concern is usually formulated. A better phrase would be to speak about
“additional” rights (i.e., rights not recognized as part of the Charter by the CJEU or as part of the Convention by
the ECtHR) insofar as they do not conflict with the rights guaranteed in Charter (or the Convention). Whether
those additional rights are “higher” in any meaningful sense of that term will likely depend on your point of view.
126 German Law Journal [Vol. 16 No. 01
Charter” to protect the “primacy, unity and effectiveness of EU law.”
77
The next paragraph
simply notes “there is no provision in the agreement envisaged to ensure . . .
coordination.”
78
It could be possible not to change anything and to argue, along the lines just discussed,
that there is no real problem with Article 53 ECHR. One could respond to the Court, then,
by explaining that the predicate (“in so far as”) to the Court’s demand of a remedy has
been rebutted. Alternatively, the Member States could, once again, go on record saying
that Article 53 ECHR does not reserve any additional rights to the Member States that the
Member States do not enjoy under the Charter. That seems like a simple fix and, again,
one that does not involve any negotiation beyond the Member States of the Union.
V. The Problem of Mutual Trust and the Ticking Bomb of Non-Accession
This leads us to one of the Court’s biggest concerns: mutual trust. Oddly, this concern does
not figure into the Advocate General’s opinion at all. At the root of this problem is a very
practical tension between the case law of the EHCR and the CJEU, especially in matters of
asylum and family law. It also reflects a profound clash between the Court’s constitutional
and the ECHR’s intergovernmental vision of the Union. As we shall see, there is a great deal
to this problem, but not necessarily in the way the CJEU sees it.
1. A Concrete Clash of Standards in Asylum Law
One pressing practical problem of mutual trust is the tension between the CJEU and the
ECtHR’s jurisprudence regarding the Dublin system on the interstate transfer of asylum
seekers to the Member State of first entry.
79
As the CJEU reiterates in Opinion 2/13, it has
77
Opinion 2/13 at para. 189. A good argument can be made that this entire paragraph is incoherent, but there is
no need to get into that here.
78
Id. at para. 190.
79
This system originated in the Dublin Convention Determining the State Responsible for Examining Applications
for Asylum Lodged in One of the Member States of the European Communities, 1997 O.J. (C 254) 1 (Aug. 19,
1997), which was replaced by Council Regulation No. 343/2003 of 18 Feb. 2003, Establishing the Criteria and
Mechanisms for Determining the Member State Responsible for Examining an Asylum Application Lodged in one
of the Member States by a Third Country National, 2003 O.J. (L 50) (“Dublin II”), and ultimately replaced by
Council and Parliament Regulation No. 604/2013 Establishing the Criteria and Mechanisms for Determining the
Member State Responsible for Examining an Application for International Protection Lodge in one of the Member
States by a Third-Country national or a Stateless Person (recast) (applicable from 1 January 2014). The asylum
process relies on a set of complementary directives: European Council Directive 2003/9/EC of 27 Jan. 2003 Laying
Down Minimum Standards for the Reception of Asylum Seekers, 2003 O.J. (L 31); European Council Directive
2004/83/EC of 29 Apr. 2004 on Minimum Standards for the Qualifications and Status of Third Country Nationals
or Stateless Persons as Refugees or as Person who Otherwise Need International Protection and the Content of
the Protection and the Content of the Protection Granted, 2004 O.J. (L 304) and corrigendum, 2005 O.J. (L204);
European Council Directive 2005/85/EC of 1 Dec. 2005 on Minimum Standards and Procedures in Member States
Granting and Withdrawing Refugee Status, 2006 O.J. (L 236), and corrigendum 2006 OJ (L 236).
2015] Opinion 2/13 – “It’s the Autonomy, Stupid!” 127
held that, in the Area of Freedom, Security, and Justice, Member States must trust each
other’s procedures for the protection of fundamental rights. In N.S., for instance, the Court
explained that an individual only has a legal claim to resist transfer to the Member State of
first entry if the sending state has evidence of “systemic deficiencies” in the receiving state
that “amount to substantial grounds for believing that the asylum seeker would face a real
risk of being subjected to inhuman or degrading treatment” in the receiving state.
80
The
Court warned, however, that not “any infringement of a fundamental right by the Member
State responsible will affect the obligations of the other Member States to comply with the
provisions of Regulation No. 343/2003.”
81
At stake, the Court added, was “the raison d’être
of the European Union . . . .”
82
The Court reiterated this view in Abdullahi.
83
The CJEU
steadfastly maintained that the “only way” in which an applicant can challenge his transfer
is by pleading “systemic deficiencies in the asylum procedure and in the conditions for the
reception of applicants for asylum in that Member State, which provide substantial
grounds for believing that the applicant for asylum would face a real risk of being
subjected to inhuman or degrading treatment within the meaning of Article 4 of the
Charter.”
84
This same standard of mutual trust and the presumption of compliance
presumably applies throughout the EU’s Area of Freedom, Security, and Justice, from
European arrest warrants to interstate transfers of child custody.
85
The ECtHR’s case law, by contrast, has been rather more solicitous of asylum seekers’
fundamental rights. In M.S.S. v. Belgium and Greece,
86
for instance, which predates the
CJEU’s N.S. decision by a few months, the ECtHR found Belgium liable under the
Convention for having transferred an asylum seeker back to Greece (which the ECtHR had
separately found to have violated Article 3 ECHR’s prohibition on inhuman or degrading
treatment).
87
The ECtHR also held that Belgium violated Article 13 ECHR, which guarantees
the right to an effective remedy, by failing to provide for a proper review of such claims.
88
80
N.S. v. Sec’y of State for the Home Dep’t, Case C-411/10, EU:C:2011:865, para. 94.
81
Id. at para. 82.
82
Id. at para. 83.
83
Abdullahi v. Bundesasylamt, Case C-394/12, EU:C:2013:813.
84
Id. at para. 62.
85
For an insightful general treatment, see Koen Lenaerts, The Principle of Mutual Recognition in the Area of
Freedom, Security and Justice, The Fourth Annual Sir Jeremy Lever Lecture, All Souls College Oxford (Jan. 30,
2015).
86
ECHR App. No. 30696/09 (Jan. 21, 2011), http://hudoc.echr.coe.int/sites/eng/pages/search.aspx?i=001-
103050#{"itemid":["001-103050"]}.
87
Id. at para. 264.
88
Id. at para. 396.
128 German Law Journal [Vol. 16 No. 01
Even though the factual situation in Greece in that case clearly amounted to widespread
failures throughout the system, nothing in the ECtHR’s judgment suggests that Belgium’s
responsibility depends on the systemic shortcomings in Greece. The way the ECtHR put it,
Article 13 ECHR demanded “independent and rigorous scrutiny of any claim that there
exist substantial grounds for fearing a real risk of treatment contrary to Article 3 [ECHR].”
89
National courts have begun to pick up on the subtle but real differences that potentially
exist between the CJEU and the ECtHR standards. The U.K. Supreme Court in E.M. v.
Secretary of State for the Home Department,
90
for instance, criticized the CJEU’s standard
for ignoring the real risk that an individual’s transfer will violate the Convention in cases
where that risk does not result from a “systemic failure” of the receiving system.
91
The U.K.
court complained:
[A]n exclusionary rule based only on systemic failures
would be arbitrary both in conception and in practice.
There is nothing intrinsically significant about a
systemic failure which marks it out as one where the
violation of fundamental rights is more grievous or
more deserving of protection. And, as a matter of
practical experience, gross violations of article 3 rights
can occur without there being any systemic failure
whatsoever.
92
The U.K. Supreme Court said the rule that “only systemic deficiencies” can constitute a
basis to resist transfer to the listed country “cannot be upheld.”
93
The only correct
approach, that Court held, was the one employed by the ECtHR, i.e. that “removal of a
person from a Member State of the Council of Europe to another country is forbidden if it
is shown that there is a real risk that the person transferred will suffer treatment contrary
to article 3 of ECHR.”
94
Shortly after the CJEU’s Abdullahi judgment and after the U.K. Supreme Court’s
intervention, the ECtHR returned to this question and the potential difference in
standards. In Tarakhel v. Switzerland, the ECtHR considered the claim by a family of asylum
89
Id. at para. 293; cf. id. at paras. 359, 365.
90
R (E.M.(Eritrea)) v. Sec. of State for the Home Dep’t, [2014] UKSC 12 (Feb. 19, 2014).
91
Id. at para. 42.
92
Id. at para. 48.
93
Id. at para. 58.
94
Id.
2015] Opinion 2/13 – “It’s the Autonomy, Stupid!” 129
seekers living in Switzerland that transferring them to Italy would violate their Convention
rights.
95
Switzerland, to which the Dublin II & III Regulations have been extended,
96
had
ordered the family’s return to Italy as the first country of entry within the Dublin system.
97
The Strasbourg Court recounts the CJEU’s case law on the Dublin system, including the
exception of “systemic risk.” It notes that the applicants themselves formulated their
claim in those terms. And yet, the ECtHR reiterates its own standard from M.S.S.
98
The
Strasbourg Court then rather ingeniously concludes:
In the case of ‘Dublin’ returns, the presumption that a
Contracting State which is also the ‘receiving’ country
will comply with Article 3 of the Convention can
therefore validly be rebutted where ‘substantial
grounds have been shown for believing’ that the
person whose return is being ordered faces a ‘real risk’
of being subjected to treatment contrary to that
provision in the receiving country.
99
Whether the word “therefore” in this statement means that the Luxembourg and
Strasbourg standards are the same or that Strasbourg’s standard trumps Luxembourg’s
standard for purposes of the Convention is anyone’s guess. Either way, it is a strong
warning signal to Luxembourg that the CJEU’s standard better comport either in words or
in practice with what Strasbourg demands or else the Dublin system violates the
Convention.
To drive the point home, the ECtHR points out that it agrees with the U.K. Supreme Court
that the “source” of the risk is immaterial and “does not exempt that State from carrying
out a thorough and individualized examination of the situation of the person
concerned.”
100
The question, then, for the ECtHR is to “ascertain whether, in view of the
overall situation with regard to the reception arrangements for asylum seekers in Italy and
95
ECHR App. No. 29217/12 (Nov. 4, 2014), http://hudoc.echr.coe.int/sites/eng/pages/search.aspx?i=001-
148070#{"itemid":["001-148070"]}.
96
Switzerland is bound by the Dublin System by virtue of the association agreement of 26 October 2004 between
the Swiss Confederation and the European Community regarding criteria and mechanisms for establishing the
State responsible for examining a request for asylum lodged in a Member State or in Switzerland (O.J. (L 53) Feb.
27, 2008). The Dublin III Regulation was passed into law by the Swiss Federal Council on 7 March 2014.
97
Tarakhel, ECHR App. No. 29217/12 at paras. 9–19.
98
Id. at paras. 100–103.
99
Id. at para. 104 (emphasis added).
100
Id. (citing the U.K. Supreme Court’s judgment in E.M.).
130 German Law Journal [Vol. 16 No. 01
the applicants’ specific situation, substantial grounds have been shown for believing that
the applicants would be at risk of treatment contrary to Article 3 if they were returned to
Italy.”
101
This sets up a clash between Strasbourg and Luxembourg on questions of mutual
trust in the Area of Freedom, Security, and Justice.
2. Brief Interlude: Three Conditions of Mutual Trust
The CJEU seems to be concerned that, once the ECHR becomes binding on the EU itself,
Member States will increasingly invoke the ECHR to disregard their EU obligations of
interstate cooperation on account of individual—as opposed to systemic—rights violations.
After all, once the ECHR binds the EU, the CJEU’s “systemic failure” standard will not only
clash with the Member States’ own “real risk” obligations under the Convention, but with
the EU’s legal obligations under the Convention as well.
102
Many might well see this as an outrageous worry.
103
How could the CJEU object to a
Member State’s refusal to send an individual into another Member State in which that
individual’s fundamental rights are at real risk of being violated? I, too, am sympathetic to
the sentiment. Indeed, it was a related outrage—though at a far higher level of moral
transgression—that tore the American Union apart in the middle of the 19
th
Century.
Abolitionist States were obligated by the Fugitive Slave Clause to deliver escaped slaves to
their “home” slave States.
104
This was a grave moral transgression. And we all know how
that ended for the American Union.
A federalism perspective illuminates how this problem that contributed to the U.S. Civil
War was similar to the mutual trust problem that troubles the European Union today.
From a federalism perspective, the deep problem of stability in both cases is a serious
mismatch between obligations of mutual trust and social reality.
I suggest that the survival of any Union that demands every component state trust and
give effect to the legal process of every other component state, depends on three
101
Id. at para. 105.
102
To be sure, some might argue that this is already the case since Article 53 CFR refers to the Convention.
Therefore, some might say that the Charter of Fundamental Rights already today requires application of the “real
risk” standard. Currently, however, the CJEU is ultimately in charge of interpreting the Charter, including the
extent to which the Charter imports the ECHR’s standards. The clash today, then, is legally a rather indirect
tension between the two standards. After accession, any difference would transform into a clear violation of the
EU’s legal obligations under the Charter.
103
Cf., e.g., Peers, supra note 3 (describing the CJEU’s resistance to Member States’ imposition of higher
fundamental rights standards as “shocking” from a human rights perspective, and especially in the light of the
Court’s ruling on mutual trust).
104
See Prigg v. Pennsylvania, 41 U.S. 539 (1842).
2015] Opinion 2/13 – “It’s the Autonomy, Stupid!” 131
interrelated conditions. First, a reasonably common set of values and similar level of
fundamental rights protection throughout the Union. Second, the Union’s ability to
remedy rights violations in component states effectively whenever they occur. And third, a
safety valve (either in primary or secondary law) for a component state to invoke
overriding policy justifications where compliance with mutual trust would otherwise rip
the Union apart.
Furthermore, there is, in my view, a hydraulic connection between these three conditions
of mutual trust: where one or more of these elements is weak, the remaining element(s)
must be correspondingly strong. For example, if there are serious divergences in
fundamental rights protections, and the Union does not have the power to step in protect
individuals, it must relax the obligations of mutual trust. At bottom, for a federal union to
survive, any legal obligation of mutual trust must be grounded in social reality, not judicial
fiat.
3. Beyond Asylum Law: The Ticking Bomb of Non-Accession
The CJEU’s demand that accession provide an exemption for Convention violations caused
by a Member State’s EU-related mutual trust obligations is rather short sighted. An
exemption would alleviate the tension between Strasbourg, Luxembourg, and Member
States in asylum law without relieving the corresponding (and far greater) tension among
these three actors that pervades fundamental/human rights in the European Union more
generally. To be sure, accession under the conditions the CJEU’s demands may help the
CJEU in its most immediate troubles regarding the concrete clash on asylum standards.
But, as I shall explain, accession on those conditions does not address the far bigger
problem that affects all of EU law.
The CJEU seems to overlook this larger problem and its relation to accession completely.
The Court also misses the fact that accession without a mutual trust exemption will help
address the larger problem. Conversely, with regard to the larger problem, the Court has
much to lose. Here, non-accession is the outcome the Court should fear. To see why, we
need to back up and briefly examine the state of fundamental/human rights protection in
the European Union.
3.1 The Wholesale/Retail Problem
Currently—in the absence of accession—the ECtHR has decided the EU cannot be sued in
Strasbourg.
105
EU actions are reviewable in Strasbourg only indirectly by holding Member
States liable either for bringing about an offensive EU measure by unanimous vote, or for
105
Cf. Bosphorus v. Ireland, ECHR App. No. 45036/98, para. 152 (June 30, 2005),
http://hudoc.echr.coe.int/sites/eng/pages/search.aspx?i=001-69564#{"itemid":["001-69564"]}.
132 German Law Journal [Vol. 16 No. 01
implementing an EU measure.
106
Currently, the ECtHR considers such challenges with
considerable deference. Member State actions are presumed to be lawful under the
Convention as long as the action is subject to a roughly equivalent standard of
fundamental rights protection within the EU.
107
As the ECtHR announced in its seminal Bosphorus judgment, however, this presumption
can be rebutted if “in the circumstances of a particular case, it is considered that the
protection of Convention rights was manifestly deficient.”
108
The ECtHR will thus still hear
challenges to a Member State’s implementation of EU law either where the EU has failed
to provide an equivalent standard of protection or where the Member State exercised
discretion in the matter, and the violation could have been avoided without disregarding
Union law.
109
The Bosphorus standard is often likened to the German Constitutional Court’s Solange
doctrine,
110
but there is an important difference between the two. Under the well-known
Solange compromise, the Member State high court will not review individual complaints
that EU actions violate fundamental rights, as long as the EU generally provides an
equivalent standard of fundamental rights protection.
111
The Solange compromise has two components. The first is rather similar to the ECtHR’s
margin of appreciation, but which we shall here simply call rough equivalence.
112
It
suggests that the standards employed by the CJEU and the ECtHR need not be identical,
but only comparable—i.e., roughly the same. The second element is what I have long liked
106
See generally, id. See also Matthews v. the United Kingdom, ECHR App. No. 24833/94 (Feb. 18, 1999); SEGI and
others v. 15 Member States (SEGI and Gestoras Pro-Amnistia v. Germany and others), ECHR App. No. 6422/02,
decision of inadmissibility of May 23, 2002.
107
See, e.g., Bosphorus, ECHR App. No. 45036/98 at para. 155.
108
Id. at para. 156.
109
This is, indeed, the posture of many of the recent asylum cases in Strasbourg. States in those cases could,
legally under EU law, have invoked the so-called “sovereignty” clause and chosen to keep the asylum seeker. The
decision to send the asylum seeker back to the state of first entry was, therefore, fully attributable to that
Member State.
110
Bundesverfassungsgericht [BVerfG – Federal Constitutional Court], Case No. 2 BvR 197/83, 73 ENTSCHEIDUNGEN
DES BUNDESVERFASSUNGSGERICHTS [BVerfGE] 339, 374, 387 (Oct. 22, 1986); cf. Bundesverfassungsgericht [BVerfG –
Federal Constitutional Court], Case No. 2 BvL 1/97, 102 ENTSCHEIDUNGEN DES BUNDESVERFASSUNGSGERICHTS [BVerfGE]
147, 164 (June 7, 2000).
111
BVerfG, Case No. 2 BvR 197/83 at para. 132.
112
For a comprehensive discussion of the margin of appreciation, see, e.g., Yutaka Arai-Takahashi, THE MARGIN OF
APPRECIATION DOCTRINE AND THE PRINCIPLE OF PROPORTIONALITY IN THE JURISPRUDENCE OF THE ECHR (2001).
2015] Opinion 2/13 – “It’s the Autonomy, Stupid!” 133
to call a wholesale/retail distinction.
113
A Member State court following the Solange
doctrine will not examine at the retail level—meaning with regard to an individual case—
whether the EU violated fundamental rights in that case. Such a court will consider only
claims that there has been a wholesale disregard for fundamental rights at the EU level—
meaning that Member States will consider fundamental rights violations of the EU only
where they exist in bulk across a range of cases. The retail business of fundamental rights
protection with regard to EU measures is the EU’s business, not that of the Member States.
As far as EU law is concerned, a Member State high court’s business under the Solange
compromise is strictly wholesale. Following this idea, for example, the German
Constitutional Court in the Solange decision itself held inadmissible a fundamental rights
claim by a mushroom exporter against Germany’s implementation of an EU export
restriction on the grounds that the EU “generally” provided fundamental rights protection
that was “essentially equivalent” to that envisaged by the German Grundgesetz
(constitution).
114
The Bosphorus standard that Strasbourg applies to a Member State’s nondiscretionary
implementation of EU actions, and the Solange standard are currently in serious tension
with one another. To be sure, Strasbourg’s standard of “equivalence” is likely to match
what we have loosely called the rough equivalence prong of the Solange compromise.
115
But the two standards differ radically on what we’ve termed the wholesale/retail
distinction.
As far as adjudicating Convention rights are concerned, the ECtHR is still firmly established
in the retail business.
116
The Bosphorus presumption can be rebutted on a case-by-case
basis.
117
But a Member State following the Solange compromise, as we have just seen, is
not. Accordingly, the German high court might well reject an individual’s fundamental
rights challenge even though the CJEU made a grave error in that individual case as long as
113
Much like a wholesaler, who does not deal with individual customers but sells goods only in bulk (or en gros),
so, too, the German high court refuses to deal with individuals’ case-by-case fundamental rights complaints,
considering only a claim that the EU has generally violated fundamental rights.
114
BVerfG, Case No. 2 BvR 197/83 at para. 132.
115
Cf. generally Arai-Takahashi, supra note 112; Howard Charles Yourow, THE MARGIN OF APPRECIATION DOCTRINE IN
THE DYNAMICS OF EUROPEAN HUMAN RIGHTS JURISPRUDENCE (1996); Steven Greer, THE MARGIN OF APPRECIATION:
INTERPRETATION AND DISCRETION UNDER THE EUROPEAN CONVENTION OF HUMAN RIGHTS (2000); J.G. Merrills, THE
DEVELOPMENT OF INTERNATIONAL LAW BY THE EUROPEAN COURT OF HUMAN RIGHTS (1993). For a recent critical assessment,
see Federico Fabbrini, The Margin of Appreciation and the Principle of Subsidiarity: A Comparison, in A FUTURE FOR
THE MARGIN OF APPRECIATION? (Mads Andenas, Eirik Bjorge & Giuseppe Bianco, eds., forthcoming, 2015), available
on SSRN at http://ssrn.com/abstract=2552542.
116
Bosphorus, ECHR App. No. 45036/98.
117
Cf., e.g., Michaud v. France, ECHR. App. No. 12323/11 (Dec. 6, 2012),
http://hudoc.echr.coe.int/sites/eng/pages/search.aspx?i=001-115377#{"itemid":["001-115377"]}.
134 German Law Journal [Vol. 16 No. 01
the CJEU has not failed to protect fundamental rights more generally. That same individual
can now go to the Strasbourg court, which will condemn Germany for implementing EU
law in that particular case because the action in that individual case manifestly disregarded
that particular applicant’s fundamental rights. Should this happen, Germany will surely
rethink the Solange compromise.
The consequences could not be more dramatic. The current tension between Strasbourg’s
retail standard and Solange’s wholesale standard threatens to unravel the entire
compromise—a core principle of judicial cooperation in the Union for the past thirty
years.
118
Even Member States who have their own version of such a compromise that does
not strictly follow the Solange model will surely take note if Germany backs away from this
foundational model and begins to review all of EU law for fundamental rights violations on
a case-by-case basis.
3.2 Why Accession Helps Solve This Problem
The Court seems to miss the fact that accession does not exacerbate the wholesale/retail
problem in asylum law or elsewhere. To the contrary, accession defuses the explosive
tension in the triangle between Luxembourg, Strasbourg, and the Member State high
courts on the wholesale/retail problem of mutual trust in asylum policy and beyond.
Once the EU accedes to the ECHR, the EU will itself be under the normal legal obligation to
protect rights according to ECHR standards.
119
According to the Draft Agreement, if a
Member State that follows the Solange compromise or the “mutual trust” obligation is
sued in Strasbourg after accession, the EU can become a co-respondent and effectively
take over the litigation. The EU can step in and take joint or, where appropriate, even full
responsibility for the violation.
118
To be sure, even Germany has since developed additional retail checks on EU law, such as ultra vires and
identity actions. See, e.g., Franz C. Mayer, Rebels Without a Cause? A Critical Analysis of the German
Constitutional Court’s OMT Reference, 15 German L.J. 111 (2014); Jürgen Bast, Don’t Act Beyond Your Powers: The
Perils and Pitfalls of the German Constitutional Court’s Ultra Vires Review, 15 German L.J. 168 (2014); Mattias
Kumm, Rebel Without a Good Cause: Karlsruhe’s Misguided Attempt to Draw the CJEU into a Game of “Chicken”
and What the CJEU Might do About It, 15 German Law Journal 204 (2014); Armin von Bogdandy and Stephan
Schill, Overcoming absolute primacy: Respect for national identity under the Lisbon Treaty, 48 Comm. Mkt. L. Rev.
1417 (2011); Mattias Kumm, The Jurisprudence of Constitutional Conflict: Supremacy Before and After the
Constitutional Treaty, 11 Eur. L. J. 262, 302–03 (2005). But the scope of these threats would likely pale in
comparison to what would be unleashed if the German Constitutional Court would step back into the retail
business of fundamental rights adjudication.
119
As Olivier de Schutter further points out, the ECtHR would have little reason to grant the EU special Bosphorustype
deference after accession. See Olivier de Schutter, Bosphorus Post-Accession: Redefining the Relationship
between the European Court of Human Rights and the Parties to the Convention, in THE EU ACCESSION TO THE ECHR
177 (Vasiliki Kosta, Nikos Skoutaris & Vassilis P. Tzevelekos eds., 2014).
2015] Opinion 2/13 – “It’s the Autonomy, Stupid!” 135
I am tempted to abuse an idiom and say this would be a “win-win-win” situation. By taking
responsibility for the violation, the EU will shield the Member State in question. Thus, the
EU will be responsible for fixing the human rights problem; Member State high courts can
cheerfully continue to defer to the CJEU under the Solange compromise; and mutual trust
will be preserved as well. To put the point somewhat colorfully: Karlsruhe can defer to
Luxembourg, and if Luxembourg fails, Brussels will step in as joint respondent and take
over full responsibility should Berlin get sued in Strasbourg.
Plain accession, then, solves the mutual trust problem the Court seems so concerned
about in the Area of Freedom, Security, and Justice, including the asylum context. There is
no need for an exemption in the Convention for Member States who follow their
obligations of mutual trust. If a Member State transfers an individual to another Member
State in compliance with mutual trust but in violation of the Convention, that Member
State can be sued in Strasbourg. The EU can then join that Member State and take full
responsibility for that action before the Strasbourg court.
Some might object that the EU might not have the power or resources to remedy the
violation in such a case. But that is the point of the hydraulic connection between what I
earlier spelled out as the “three conditions of mutual trust.” Where the EU has the power
to intervene and remedy the rights violation,
120
it will be bound by the Convention to do
so. Where the EU lacks the power or the resources, it will be forced by the Convention to
open what we have termed the necessary “safety valve” in the principle of mutual trust. In
those cases, the EU, whether by CJEU order or legislation, will need to allow the Member
States to check one another. This reserves to the EU, as opposed to Strasbourg, the
calibration of the three elements. Luxembourg, not Strasbourg, determines the nature and
depth of mutual trust. And yet Strasbourg can make sure that rights don’t get dismissed
along the way.
With respect to the Dublin System, this is quite easily done. Under the so-called
“sovereignty” clause, Member States may already choose to allow an asylum seeker to
remain within their territory and process the application without sending the individual
back to the Member State of first entry.
121
The EU could rather simply give in and adopt
120
For arguments in favor of expansive EU powers on this score, see, e.g., K.L. Scheppele, EU Commission v.
Hungary: The Case for the “Systemic Infringement Action”, VERFBLOG, (Nov. 22, 2013),
http://www.verfassungsblog.de/en/eu-commission-v-hungary-the-case-for-the-systemic-infringement-action/
(suggesting EU systemic infringement action against Member States). Compare further Daniel Halberstam,
Constitutional Heterarchy: The Centrality of Conflict in the European Union and the United States, in RULING THE
WORLD? CONSTITUTIONALISM, INTERNATIONAL LAW, AND GLOBAL GOVERNANCE 326, 351–353 (Jeffrey L. Dunoff & Joel P.
Trachtman eds., 2009) (suggesting reverse-Solange review to protect citizens’ fundamental rights), with Armin
von Bogdandy et al, Reverse Solange–Protecting the essence of fundamental rights against EU Member States, 49
Comm. Mkt. L. Rev. 489 (2012) (fleshing out same).
121
Regulation No. 343/2003, art. 3(2). Indeed, this is the posture of the current cases in Strasbourg. In the cases
discussed earlier in the text, Strasbourg held that Member States bore full responsibility for the transfer because
136 German Law Journal [Vol. 16 No. 01
the ECtHR’s standard, allowing individuals to claim a “real risk” of inhuman and degrading
treatment in the receiving state, regardless of any “systemic” dimension of the problem.
122
The EU legislature could adopt that provision as law, or the CJEU (upon accession) could
interpret the Regulation’s sovereignty clause in conformity with the EU’s legal obligations
under the EHCR.
However the balance is achieved, the result will be that the EU’s legal control of the degree
of mutual trust its Member States must show each other and the EU will be preserved.
Human rights under the Convention will be preserved. And the Solange compromise that
forms an important element of constitutional pluralism in the EU will be preserved as well.
4. How Do We Get There From Here?
The CJEU, then, has identified a substantive problem that seems to be far greater than
even the Court itself may have realized. As a result, the Court’s proposed remedy asks for
both too much and too little.
Opinion 2/13 seems to be pushing the analogy to a traditional federal state too hard. By
asking for an express exemption for Member States’ Convention violations caused by EU
law’s mutual confidence obligations, the CJEU is trying to mimic the existence of a federal
state in international law. After all, where component units of a federal state implement
federal law in violation of a treaty the federation itself has signed, international law holds
the center—not the component units of government—responsible.
123
But simply
transferring this rule to the EU for cases involving the Area of Freedom, Security, and
Justice does both too much and too little. First, it fails to account for the many ways in
which the EU is not an integrated state, especially in that its Member States are generally
and broadly full-fledged international states. Second, the proposed remedy seems
principally focused on the Area of Freedom, Security, and Justice and on the relations
among the Member States themselves, ignoring the larger threat to vertical judicial
cooperation under the Solange compromise.
they could have allowed the individual applicant to remain in their territory pursuant to the “sovereignty” clause.
The Strasbourg court then held that the Member State has an obligation under the Convention to exercise this
option as a way to protect the individual’s rights under the Convention. See, e.g., Tarakhel, ECHR App. No.
29217/12.
122
M.S.S. v. Belgium and Greece, ECHR App. No. 30696/09 (Jan. 21, 2011), para. 293,
http://hudoc.echr.coe.int/sites/eng/pages/search.aspx?i=001-103050#{"itemid":["001-103050"]}; cf. Tarakhel,
ECHR App. No. 29217/12 at paras. 359, 365.
123
See, e.g., Draft Articles on Responsibility of States for Internationally Wrongful Acts, with Commentaries, [2001]
2 Y.B. Int’l L. Comm’n 2, ¶ 77, U.N. Doc. A/CN.4/SER.A/2001/Add.1 (Part 2).
2015] Opinion 2/13 – “It’s the Autonomy, Stupid!” 137
Fortunately, the solution to this conundrum may already be contained in changes to the
Draft Agreement discussed earlier: (1) eliminating the power of the ECtHR to second guess
the EU’s bid to become a co-respondent, and (2) eliminating the ECtHR’s power to second
guess the EU’s view on joint versus sole responsibility. As long as those two changes are
made (along the lines discussed earlier), the Commission can present to the Court that it
has heard the Court’s concerns, that changes have been made, and that Member States’
obligations of mutual trust will be shielded from ECtHR interference through the corespondent
mechanism.
VI. CFSP Jurisdiction and the “Consolidating Function” of the Court
The final part of Opinion 2/13 has been criticized as “mind-boggling,” as “politics of the
playground,” and as giving rise to “a moral duty to reject the EU’s accession to the
ECHR.”
124
The Advocate General, too, finds no fault in granting jurisdiction over CFSPrelated
human rights claims to Strasbourg even in cases where the CJEU has no jurisdiction
over the matter. If there is no supranational governance in CFEU matters, as the AG says, is
there any privileged CJEU role to protect? And isn’t some review better than no review at
all? The Court’s decision seems like a naked power grab.
Not so fast. You do not have to be naïve about judicial motivation to see that there is more
to the Court’s opinion than playground politics. Once again, if one is exclusively focused on
international human rights regimes, it may indeed be difficult to fathom what objection
one could possibly have to bringing another actor under the umbrella of the ECtHR. The
goal of rights protection, after all, is laudable and righteous.
But the human rights perspective does not negate the fact that, from a constitutional
perspective of the Union, there may nonetheless be legitimate concerns about the Draft
Agreement in the area of CFSP. This will take a few moments to explain. But in a nutshell,
the Court’s constitutional concern can be summarized as this: The CJEU must be allowed to
play a “consolidating function” if domestically justiciable claims that European Union law
violates fundamental/human rights are brought before an international court.
1. The Court’s Constitutional Claim
Let’s start with the Court’s concerns. A deceptively simple constitutional idea lies at the
beginning of this argument: there is only one European Union. To be sure, shadows of the
Maastricht Treaty’s pillar system remain even after passage of the Lisbon Treaty. For
example, CFSP matters often function according to special rules, including unanimity voting
124
Steve Peers, The CJEU and the EU’s Accession to the ECHR: A Clear and Present Danger to Human Rights
Protection, EU LAW ANALYSIS BLOG (Dec. 18, 2014), http://eulawanalysis.blogspot.it/2014/12/the-cjeu-and-eus-
accession-to-echr.html?utm_source=Weekly+Legal+Update&utm_campaign=b9719ea1b3-
WLU_19_12_2014&utm_medium=email&utm_term=0_7176f0fc3d-b9719ea1b3-422285509.
138 German Law Journal [Vol. 16 No. 01
rules,
125
lack of legislative power,
126
and—most important for present purposes—the lack
of CJEU jurisdiction over certain matters in this domain.
127
CFSP actions therefore sure look
like they still fall into a rather distinct category from the rest of Union activity. But still,
regardless of voting rules and judicial review, there is only one European legal order.
Even when we had the Community on the one hand, and the Union on the other, the Court
pushed for a reading that integrated important background principles among the two.
Read only Pupino,
128
where the Court roundly rejected Member States’ arguments that
Justice and Home Affairs (“JHA”) “framework decisions and Community directives are
completely different and separate sources of law,” and that Member State courts did not
have a duty to interpret national law in conformity with framework decisions because of
the “inter-governmental nature of cooperation between Member States in the context of
Title VI of the Treaty on European Union.”
129
Despite unanimity rules and the lack of direct
effect under then-Article 24 TEU, and the optional nature of JHA reference actions under
then-Article 35 TEU, the Court held that the usual Community obligations of loyalty and
conforming interpretation applied to the Union’s JHA measures as well. To be sure, that
was JHA, and this is CFSP where the Court may be said to have even less involvement.
130
Nonetheless, with the formal merger of the Community and the Union after Lisbon, there
is only one EU legal order.
That is the big constitutional claim. Although specific elements of the legal order might be
adjusted in different areas of functioning, the strong presumption is that the great
background principles of the European Union’s legal order should remain the same.
131
As a
matter of interpretation, then, in an area such as CFSP, we begin with constitutional
principles and ask how these might be minimally altered by express exceptions of the
Treaty. We no longer think of CFSP as a separate legal order broadly governed by its own
general intergovernmental principles; nor do we take the thin governance provisions of
CFSP as suggesting a more intergovernmental vision of the Union as a whole. This is the
125
See, e.g., TEU arts. 22, 24, 31(1).
126
See, e.g., TEU arts. 24(1), 31(1).
127
TEU art. 24(1), TFEU art. 275.
128
Judgment Criminal Proceedings Against Maria Pupino, Case C-105/03, EU:C:2005:386.
129
Id. at paras. 25–26
130
Cf. Christophe Hillion & Ramses Wessel, Restraining External Competences of EU Member States under CFSP, in
EU FOREIGN RELATIONS LAW: CONSTITUTIONAL FUNDAMENTALS 79 (Marise Cremona & Bruno de Witte eds., 2008).
131
Cf. Christophe Hillion, A Powerless Court? The European Court of Justice and the Common Foreign and Security
Policy, in THE EUROPEAN COURT OF JUSTICE AND EXTERNAL RELATIONS LAW: CONSTITUTIONAL CHALLENGES 47, 49–51 (Marise
Cremona & Anne Thies, eds., 2014).
2015] Opinion 2/13 – “It’s the Autonomy, Stupid!” 139
constitutional approach that serves as the basis for the complaint about the Draft
Agreement.
2. The Accession Picture: Before and After
Currently, the CJEU lacks jurisdiction over certain CFSP measures.
132
To the extent
jurisdiction over CFSP matters has not been delegated to the CJEU, Member States and
their courts retain jurisdiction.
133
As a result, Member State courts currently have the final
say on some matters of EU law (even where the EU is a party to the dispute).
134
Accordingly, until the CJEU clarifies the full extent of its jurisdiction under Article 275 TFEU
(which it notably refused to do in Opinion 2/13), Member State high courts have the final
say on the interpretation of Charter rights at least in some CFSP matters.
Currently, then, in the absence of CJEU jurisdiction, there is no single EU institution to
harmonize potentially conflicting interpretations of EU law in this area. This is more than
just “regrettable.”
135
From a constitutional perspective, it is deeply problematic, especially
in an area presumably vital to the international security of the Union. Recall Justice
Holmes’ famous dictum: “I do not think the United States would come to an end if we lost
our power to declare an Act of Congress void. I do think the Union would be imperiled if
we could not make that declaration as to the laws of the several states.”
136
Fully
decentralized rights adjudication can create similar challenges as unreviewable component
state legislation.
137
One can well affirm the value of preserving Member State courts’
competing power to interpret the rules of the system in a constellation of pluralism, and
still see that denying the CJEU all power to adjudicate the law of the Union is in serious
tension with the constitutional idea of a Union. If component state courts retain exclusive
authority to determine the legality of Union law, the Union may well be in peril.
Nonetheless, the current situation does not violate the EU’s constitutional order, or render
it a sham. First, the Treaty expressly limits the CJEU’s jurisdiction over CFSP measures along
132
See TEU art. 24 (1); TFEU art. 275.
133
See generally Opinion of Advocate General Mengozzi, Gestoras Pro Amnistía and Others v. Council, C‑ 354/04
P and C‑ 355/04 P, EU:C:2006:667 (Oct. 26, 2006) (albeit predating the Treaty of Lisbon).
134
TFEU art. 274.
135
Opinion 2/13 at para. 101.
136
Oliver Wendell Holmes, COLLECTED LEGAL PAPERS, 295–296 (1921).
137
The 19th
Century U.S. episodes of nullification and interposition were in part based not on unreviewable
component state laws, but on component states’ assertion of their authority to interpret the national
constitution. See H. Jefferson Powell, Joseph Story’s Commentaries on the Constitution: A Belated Review, 94 Yale
L.J. 1285, 1292 (1985).
140 German Law Journal [Vol. 16 No. 01
the lines just discussed.
138
Second, as the Treaty fairly suggests in Article 19(1) TEU,
Member State courts are, after all, courts of the Union.
139
Accordingly, when Member
State courts interpret EU law in the course of adjudicating CFSP matters, they have an
obligation to view themselves as engaged in a common enterprise of collective governance
with all the obligations of mutual fidelity that their own constitutions, as well as EU
primary law and general legal principles, demand.
Today, judging by Strasbourg’s UN-related case law, the ECtHR can adjudicate CFSP-related
action only indirectly, i.e., only insofar as the action is at least partly attributable to a
Member State. If a Member State is partly responsible, the ECtHR will charge that Member
State with liability under the Convention for the Member State’s own EU-related CFSP
activity.
140
Conversely, following the current line of jurisprudence, once the action is
exclusively attributable to the EU, the Strasbourg court would not have jurisdiction
because the EU is not a Contracting Party to the ECHR.
141
To be sure, this potentially
creates a troublesome accountability gap (albeit not one of the CJEU’s making) regarding
the administration of the ECHR’s human rights regime. Focusing on the Court’s
constitutional concerns about the EU’s legal order, however, we observe that, today,
Strasbourg reviews only the conventionality of Member State actions. The legality of CFSP
actions under EU law, by contrast, is currently determined exclusively by courts of the
Union, i.e., Member State courts.
Once the EU becomes a Contracting Party of the Convention, Strasbourg will gain authority
to adjudicate the ECHR compatibility of the EU’s CFSP measure itself. Some might believe
this ought not to be problematic, even from the perspective of EU law. After all, one might
say, Strasbourg would be adjudicating only the “conventionality” of the CFSP measure, and
Strasbourg’s judgments do not have immediate legal effect in Contracting States’ legal
systems. And so, then as now, Strasbourg would not be adjudicating EU law as such.
This final line of argument may be technically accurate, but it risks ignoring the substantial
practical effect that accession on the DA’s terms may have on the constitutional system of
the Union.
138
TFEU art. 275.
139
TEU art. 19(1).
140
See, e.g., Al Jedda v. United Kingdom, ECHR App. No. 27021/08 (July 7, 2011). Cf. Marko Milanovic, Al-Skeini
and Al-Jedda in Strasbourg, 23 EUR. J. INT’L L. 121 (2012).
141
Behrami and Behrami v. France and Saramati v. France, Germany and Norway, ECHR App. Nos. 71412/01,
78166/01 (May 2, 2007).
2015] Opinion 2/13 – “It’s the Autonomy, Stupid!” 141
3. The Practical Fusion of Legality and Conventionality
Accession on terms of the Draft Agreement would have effectively turned the Strasbourg
Court into the constitutional court of the Union—at least for certain matters. Recall that
the ECHR is integrated into EU law as a binding international agreement; that the EU will
be deemed to have implemented the ECHR by virtue of Article 52(3) FRC; and that the
ECHR will therefore become fully binding on EU institutions and the Member States with
priority over secondary EU law.
142
After accession on the Draft Agreement’s terms,
Member States (and their courts) would likely have taken the Strasbourg court as the
authoritative guide on the fundamental rights compatibility of the CFSP measures Member
States implement, at least whenever the CJEU could not say otherwise. As a practical
matter, Member State courts would likely have taken the ECtHR as their guide on the
legality of CFSP matters under EU law despite the fact that this authority formally was
beyond the ECtHR’s jurisdiction.
Protocol 16 would have made this outcome more likely, as it inserts the Strasbourg court
institutionally into those signatory states’ process of adjudication. This allows Member
State high courts to be in direct communication with Strasbourg on matters of EU law
without the buffer of the CJEU as a consolidating authority. From the perspective of the
Union as a whole, this risks institutionally disaggregating the Union, much as Article 177
EEC (now 267 TFEU) so brilliantly did with the Treaty of Rome.
In some respects, the disaggregating effect of Protocol 16 would likely have been even
more extreme than it was in the case of Article 177 EEC. The EU’s preliminary reference
action has always operated alongside, and in competition with, existing Member State high
courts. That, once again, was and is the stuff of constitutional pluralism. As envisioned
under the Draft Agreement’s accession conditions, by contrast, Protocol 16 would have
operated an advisory opinion system in certain CFSP matters without the EU’s own high
court anywhere in sight.
To be sure, Strasbourg judgments do not claim supremacy or direct applicability; and
Strasbourg would formally adjudicate only the compatibility of CFSP measures with the
ECHR as a matter of ECHR Treaty law. Nonetheless, because the ECHR itself would be
binding on the EU with all the resulting legal effects within the EU’s legal order, Member
State high courts might well have taken Strasbourg’s word on compatibility with the
Convention as seriously as they take any CJEU judgment on the legality of a CFSP measure
under EU law. In practical terms, then, with regard to fundamental/human rights,
Strasbourg’s conventionality review might well have operated as the EU’s legality review.
142
See Judgment in Haegeman, Case 181/73 at para. 5. For brief reflections, see Bruno de Witte, Beyond the
Accession Agreement: Five Items for the European Union’s Human Rights Agenda, in THE EU ACCESSION TO THE ECHR
349 (Vasiliki Kosta, Nikos Skoutaris & Vassilis P. Tzevelokos eds., 2014).
142 German Law Journal [Vol. 16 No. 01
Member State courts could surely still have insisted on their own internal reading of the
Charter (or in extremis perhaps even of the Convention), by stepping into the shoes of an
absent CJEU. But likely they wouldn’t have done so. Given the multiplicity of Member State
high courts, accession on terms of the Draft Agreement would have placed Strasbourg into
the position of a singular, uniquely situated, and legally privileged harmonizing voice
adjudicating a binding fundamental rights regime that is an integral part of EU law.
Add to this that Member State systems are increasingly treating obedience to EU law and
ECHR law similarly, often by virtue of domestic constitutional command.
143
Accession on
the Draft Agreement’s terms, again, would have meant that EU law itself commands
adherence to the ECHR with regard to all secondary EU law obligations (and possibly
more).
144
One must remember that, despite Article 19(1) TFEU, Member State judges are
principally trained at home and steeped in their domestic judicial bureaucracy. It might
well have been hard for them not to look to the ECtHR for complete and final resolution of
any conflict between EU law and the ECHR. Without the CJEU in sight, Member State court
judges may well have conceived of EU law and ECHR law as fused and fungible. After all,
for Member State courts both sets of norms are constitutionally imported from Europe
beyond the nation state.
In summary, Member State courts might well have taken Strasbourg’s decisions on the
conventionality of CFSP mandates beyond the purview of the CJEU as a final decision on
the legality of the action under EU law.
If this is right, accession on Draft Agreement terms would have meant that in CFSP matters
beyond Luxembourg’s jurisdiction, Strasbourg would effectively have become the
European Union’s constitutional court. To be sure, this assessment includes a good deal of
(contestable) predictions about judicial behavior. But the CJEU gets only one shot at
identifying and preventing the potential problem now. Once the Court gives the Draft
Agreement a pass, by contrast, accession is for the ages.
4. The Consolidating Function of the Court
The absence of CJEU jurisdiction in the face of EU accession would have denied the Court
all authority to help “consolidate” domestic jurisprudence on matters that result in the
143
Cf. Giuseppe Martinico, Two Worlds (Still) Apart? ECHR and EU Law before National Judges, in THE EU ACCESSION
TO THE ECHR 141 (Vasiliki Kosta, Nikos Skoutaris & Vassilis P. Tzevelokos eds., 2014).
144
Given the ECHR’s substance as a fundamental rights catalogue and its deep structural incorporation by
reference and, after accession, as binding international agreement into the core of fundamental rights protection
in the EU, the ECHR will, after accession, undoubtedly have quite more than the usual legal effect of an ordinary
international agreement to which the EU is a party.
2015] Opinion 2/13 – “It’s the Autonomy, Stupid!” 143
international liability of the Union as a whole. Given the lack of CJEU jurisdiction over
certain CFSP matters, accession on these terms places the EU on quite a different—and
rather disfavored—footing, as compared to all other Contracting Parties. All the other
Contracting Parties to the Convention have the benefit of a consolidating domestic
judiciary to harmonize interpretation and judicial review of domestic law before a case
against them proceeds to Strasbourg. The EU does not. This is more than playground
politics.
As already discussed, this lack of a domestic consolidating jurisdiction may contribute to
the fusion of conventionality and legality review as Member State courts deal severally
with Strasbourg. But there are two additional concrete consequences of denying
Luxembourg the authority to consolidate domestic jurisprudence after accession.
First, given the multiplicity of Member State court voices and the absence of the CJEU, the
ECtHR will likely wind up as a kind of soft interpreter of the underlying CFSP law as well. In
light of the absence of any harmonizing voice at the EU level, if different Member States
find themselves before the ECtHR with conflicting views of what an underlying CFSP
measure demands, the Strasbourg court will have great leeway in basing its decision on
what it takes to be the most plausible of the various interpretations of EU law. Given
Strasbourg’s privileged position in this arrangement and the absence of the CJEU,
Strasbourg’s version of EU law is likely to seep back into the jurisprudence of Member
State courts. The consolidating function of EU law in the area of CFSP will have been
farmed out to Strasbourg.
Second, the EU as a whole may incur international legal responsibility as a result of a
violation that could have been avoided if only the CJEU would have had jurisdiction. What
is more, the entire EU may be held in violation of the Convention in Strasbourg on account
of the decision of a single Member State high court, even if all other Member State high
courts (and the CJEU) would have decided the matter differently and kept the Union (or
placed the Union) in compliance with the Convention.
It is one thing expressly to preclude CJEU consolidation of the law in certain CFSP issues, as
is the case today. After all, the effects of a Member State high court decision on CFSP
issues could be (and likely should be) limited to the territorial jurisdiction of that particular
high court. After accession, however, the decisions of any given high court would have
entailed direct international responsibility of the Union as a whole without the possibility
of any recourse to a consolidating court within the Union. That would have been an
entirely new and problematic consequence for the Union as the result of accession on the
Draft Agreement’s terms.
Third, the importance of the consolidating function of a domestic high court is
corroborated by the double standard that would have implicitly been built into Protocol 16
vis-à-vis the European Union after accession under the Draft Agreement. Notice that the
144 German Law Journal [Vol. 16 No. 01
Member States only allow designated “highest” courts to ask Strasbourg for advisory
opinions.
145
Under Protocol 16 Member States would not allow their regional high courts,
for instance, to contact Strasbourg directly. As far as each Member State is concerned,
then, with or without Protocol 16, the domestic consolidation of jurisprudence is fully
preserved in interacting with Strasbourg. In the case of the EU, by contrast, accession
according to the Draft Agreement would have meant that all Member State high courts
could ring up Strasbourg while Luxembourg is shut out of the conversation.
5. Now What?
Let us return for a moment to the current “human rights gap” in CFSP matters.
Commentators have condemned Opinion 2/13 by saying that accession on the terms of the
CJEU’s opinion is not worth pursuing.
146
But, the human rights gap and the Court’s
concerns could both be accommodated rather comfortably by granting the CJEU
jurisdiction over all matters that could, upon accession, come before the Strasbourg court.
Everyone would win as a result.
This change would not require any involvement of non-EU members of the Council of
Europe, which means, most concretely, that it would not involve reopening negotiations
with Russia. From the perspective of human rights, moreover, nothing is gained by
allowing the ECtHR to adjudicate Convention violations of CFSP measures while
withholding from the CJEU jurisdiction to review those actions for their compatibility with
the Charter and, as a matter of EU law, the Convention. And if the only explanation for this
particular outcome is realpolitik, then that is a rather weak justification for a constitutional
court such as the CJEU to accept.
F. Conclusion
It is tempting to speculate about the motives of the various actors involved in this intricate
problem. I sometimes like to think of it as administering a kind of “Rorschach test,” the test
that asks what people see in an inkblot. When human rights champions see the CJEU, they
see an economically minded court that puts fundamental rights second. When human
rights champions see Strasbourg, by contrast, they see a court single-mindedly devoted to
their cause. When Member State governments see the CJEU, they see an institution that
takes a mile when you give it an inch. When Member States see the Strasbourg court, by
contrast, they see a far more respectful court, a court that is more tethered to its
signatories, and (as far as adjudicating military interventions is concerned) they see the
145
See Protocol No. 16 to the European Convention for the Protection of Human Rights and Fundamental
Freedoms, Feb. 2013, C.E.T.S. No. 214, arts. 1, 10.
146
See, e.g., Douglas-Scott, supra note 3 (“in the light of the ECJ Opinion, those who value human rights no longer
have any reason to pursue EU accession to the ECHR.”).
2015] Opinion 2/13 – “It’s the Autonomy, Stupid!” 145
devil they know. As for what the CJEU and the ECtHR see when they look at each other, the
negative images of that sibling rivalry need not be spelled out here. On the positive side,
however, they see each other as partners in keeping their signatories honest and wedded
to principle, not politics.
A cynical view sees all these actors as interested only in their own self-aggrandizement.
That is always a quick and very easy judgment to make, including when it comes to the
Court of Justice of the European Union. But, as I have tried to argue, that judgment in this
case is too quick, and far too easy. If we diligently reconstruct the Court’s various concerns
with a modicum of charity, we see that the Court has a number of valid objections about
the EU’s accession to the ECHR on the terms of the Draft Agreement.
To understand the substance of these concerns, however, we need to take the Court’s
perspective as the supreme and constitutional court of a federal-type Union (even as the
Court stands in a productive pluralist rivalry to other high courts in and of the Union).
Opinion 2/13 is, moreover, an attempt to flesh out the Union’s multi-level nature in the
world of international law. To be sure, discussions about mixed agreements and vertical
arrangements in European foreign affairs have been around for as long as the ERTA
147
case.
But once Kadi took the position that the EU is an autonomous constitutional order,
vindicating the constitutional principles of the Union’s federal-type structure in the
international arena took on added importance.
148
We are still feeling our way about in
what continues to be an ever-fresh frontier: the intersection of the Union’s multi-level
system of plural constitutionalism and the world of international law.
One must further bear in mind that this particular international agreement is like no other.
Due to the EU’s special embrace of fundamental rights, of international law generally, and
of the ECHR (via Article 53 of the Charter) in particular, the agreement goes to the very
core of the Union’s legality and legitimacy. In many ways, it is the EU’s very openness to
fundamental/human rights in and through international law that lies at the root of the
problems surrounding accession. As a matter of substance, the EU’s legal order is already
joined at the hip with the content of the Convention. With the proposed accession, the
Union will become both normatively and institutionally fused with the Convention as well.
This Article comprehensively analyzed each of the Court’s many concerns, and
reconstructed all but one—the concern about Article 53 of the Convention. As we have
seen, one of the largest concerns of the Court, that of preserving Member States’
obligations of good faith toward one another, turns out to be even larger than the Court
147
Comm’n of the European Cmtys. v. Council of the European Cmty., Case C-22/70 (21 Mar. 1971). For an early
seminal discussion, see Joseph Weiler, The External Relations of Non-Unitary Actors: Mixity and the Federal
Principle, in MIXED AGREEMENTS (David O’Keefe & Henry G. Schermers eds., 1983).
148
Kadi, Case C-402/05 P, C-415/05 P.
146 German Law Journal [Vol. 16 No. 01
imagined. Indeed, the discussion revealed that non-accession to the Convention presents a
ticking bomb for the Union’s legal order as a whole. The Court’s other large concern, that
about accession in light of the CJEU’s lack of jurisdiction over certain CFSP matters,
required a good deal of unpacking. But even here we have seen that a reasonably
charitable look at the Court’s opinion reveals real concerns about important constitutional
principles. Only after these principles are protected, can we turn to the pluralist
engagement with Strasbourg.
All of the Court’s concerns, as I have attempted to show, are solvable either unilaterally or
by a minimum of changes affecting other parties. But this does not mean accommodating
every one of the Court’s imagined concerns or proposed remedies just the way the Court
may have hoped. And it does not preclude a vigorous relationship of cooperation and
challenge between Luxembourg and Strasbourg when we are done. Revisions can be made
to accommodate the valid concerns in good faith while still doing justice to the Convention
and to the idea of pluralist contestation. The EU’s political institutions and the Member
States can and should push back where the Court’s concern is misguided or the Court’s
seemingly preferred remedy goes too far. Most wisely, of course, they would nonetheless
take the position that they are addressing all of the Court’s concerns either explicitly or
implicitly in the various changes to accession they make.
Once the EU’s political institutions and the Member States accommodate the Court in
good faith along the lines outlined here, even if not necessarily following the most
restrictive interpretation of Opinion 2/13 possible, they will have done their part. It would
then be wise to take the revisions back to the Court for approval before signing the
agreement. But they could, of course, always decide that, this time around, they will do
what President Washington was required to do in 1793, and forge ahead on their own
advice.