Debates about human rights have been instrumental in forming the EU legal order as we know it today. In the Solange saga, judges in Luxembourg went some length to ‘discover’ human rights as unwritten general principles in order to fend off challenges to the primacy of Union law by the Bundesverfassungsgericht. At the time, the case-law often limited itself to abstract declarations of intent without fully giving teeth to human rights in practice. Only recently, the newly drafted Charter of Fundamental Rights seems to have motivated the European Court of Justice (CJEU) in Luxembourg to change course. Ever since the Lisbon Treaty came into force, the judges actively explore the human rights dimension of most cases. Against this background, the forthcoming accession of the EU to the European Convention on Human Rights (ECHR) can be described as the culmination of a long process which will finally embed the EU into the pan-European human rights architecture. In future, the EU will be officially bound by the ECHR in the same way as Member States and be subject to the jurisdiction of the European Court of Human Rights (ECtHR) in Strasbourg.
I applaud the EU’s accession and I am optimistic that it will foster the credibility of both the Convention system and the EU legal order. But accession is not only a grand political project. It has a technical dimension which raises a number of delicate legal issues. The sheer complexity of the undertaking is illustrated by the voluminous Draft Accession Treaty which comprises 12 articles together with an extensive explanatory report of no less than 20 pages. It completes 3 years of protracted negotiations during which the Commission had to convince, among others, Russia and Turkey that the EU’s byzantine legal and institutional structure required special rules and procedures. It was quite successful – and yet the Draft Accession Agreement provokes a couple of questions, bringing us back to the original challenges to the primacy of Union law, which the CJEU has always been eager to deter. It might do so again: just before the summer recess, the European Commission referred the matter to the CJEU in Luxembourg, in Opinion 2/13 whether the Draft Accession Agreement falls foul of the EU Treaties.
One Dispute – Two (Potential) Respondents
The hardest nut which negotiators have had to crack emanates from the cooperative nature of European federalism, with the EU focusing on legislative harmonisation which is to be implemented by the Member States. As a result, most citizens and companies are confronted with EU law indirectly, when it is applied by state authorities or national courts. What appears at first sight to be a national measure, is often a dispute which revolves around matters of EU law. This had left negotiators with a simple question: against whom should individuals address their complaints in such cases?
The solution of the Draft Accession Agreement is simple: whenever state authorities implement EU law, Member States will be primary respondents (Article 1). This rule of attribution extends to situations of full agency where Member States have little or no discretion in the implementation of EU law, such as a seizure of an aircraft by Irish authorities in line with a mandatory EU sanctions regime – a situation in which the ECtHR would nowadays decline scrutiny, since a Member State ‘does no more than implement legal obligations flowing from [EU membership].’ Different rules are meant to apply, however, in future: EU law may be the key reason for a complaint in Strasbourg and nonetheless national authorities have been designated as primary respondents.
In order to give the EU an opportunity to defend its rules, negotiators propose, instead, to introduce the so-called ‘co-respondent mechanism’: EU institutions ‘may’ join any complaint involving supranational law. In this case, the European Union becomes a full party to the dispute with equal rights and is jointly responsible, together with the Member State, to abide by the final judgment (Article 3). If the mechanism functioned well in practice, it would present a pragmatic procedure, although purists of the Community method may grumble that Member States are assumed to be ‘jointly liable’ for a violation, such as the seizure of an aircraft, for which supranational law bears sole responsibility. Notwithstanding such detail, it seems to me that while the co-respondent mechanism addresses some intricacies of the EU’s federal system, it fails to circumnavigate all shallows.
Pitfalls of Voluntary Co-Respondence
Any lawyer specialising in EU law knows that challenges to human rights in Union law rarely result in the annulment of legislation (ongoing debates about the legality of the data retention directive may be an example). In most cases, the interaction is more indirect whenever the Court interprets legislation in accordance with the Charter and upholds human rights by means of consistent interpretation. Arguably, this line of reasoning defines the newly found prominence of human rights in Luxembourg. Human rights arguments are linked to questions of statutory interpretation. Regular judicial hermeneutics and the promotion of human rights go hand in hand.
By contrast, the Draft Accession Agreement concentrates on validity disputes. The trigger for the co-respondent mechanism relates to disputes which ‘call into question the compatibility’ of Union law with the European Convention, ‘notably where [a] violation [of the ECHR by a Member State] could have been avoided only by disregarding an obligation under European Union law.’ Such hard cases are certainly fascinating, but judges in Luxembourg know too well that such Foto Frost-style disputes are not the bread and butter of human rights case law. In my view, the threshold has been defined too narrowly. The EU should not only be involved, whenever the validity of EU legislation is at stake.
To be sure, the Draft Accession Agreement invites the ECtHR to handle the trigger for the co-respondent mechanism generously. It shall assess only whether its fulfilment ‘is plausible’ and ‘appears’ to be met. But despite this generous wording, the ECtHR’s retains the right to reject the EU’s participation. Moreover, the Draft Accession Agreement assigns the risk of non-participation to the European Union. If the Commission’s ex ante-screening fails to identify relevant cases involving the implementation of Union law among the 21,189 complaints brought against 28 EU Member States in 2012, the EU loses out on the option to be a co-respondent. Member States do not bear that risk.
Prioritising the Autonomy of EU Law …
One of the core motivations behind the EU delegation’s agreement with the co-respondent mechanism is it’s concerns about the autonomy of the EU legal order, since the Court of Justice had found the original draft agreement establishing the European Economic Area to be incompatible with the present EU Treaties precisely because the intended EEA Court would have had jurisdiction to rule on the division of competences between the EU and its Member States. The Draft Agreement on Accession to the ECHR successfully circumnavigates this cliff by neatly attributing any domestic application of EU law to Member States. Doing so protects the CJEU’s interpretative autonomy regarding the division of competences, but it arguably poses challenges to the primacy of EU law.
Let me be clear: I do not object to the ECtHR overseeing EU law. Quite the contrary, full and unconditional scrutiny of Union law by the ECtHR is overdue. I am concerned, however, about indirect challenges to the uniform interpretation and effective application of EU law. Why? The co-respondent mechanism may have the (unintended) result of side-lining the EU law dimension in national courts. The danger is not ECtHR control of the CJEU, but ECtHR decisions instead of prior CJEU involvement. Let me explain.
… to the Detriment of Primacy
The EU’s system of cooperative federalism functions extremely well, because the CJEU has developed mechanisms to ensure the uniform interpretation and effective application of Union law by national courts. The procedural backbone is the preliminary reference procedure, which the Court protects eagerly by threatening damages in cases of non-compliance, prohibiting rules which would render prior involvement of constitutional courts obligatory or preventing the unified patent litigation system – thereby safeguarding primacy against barriers to the CJEU’s direct dialogue with national judges.
If the Draft Accession Agreement went ahead, national judges would in future have an excuse not to refer human rights disputes involving Union law to Luxembourg. They could simply defer the matter, since any activation of the co-respondent procedure entails optional CJEU involvement. Indeed, the draft text explicitly foresees that Luxembourg shall be engaged, if it has not been already (Article 3.6). Of course, such involvement depends on EU participation. If the Commission fails to activate the co-respondent mechanism or if, hypothetically, the ECtHR rejects EU participation, Luxembourg will not be heard. Again, the risk of non-participation in the co-respondent mechanism which entails the option of prior CJEU involvement lies with the European Union.
To make things worse, national courts of last instance will in future have an additional motivation not to refer human rights disputes to Luxembourg. Protocol No. 16 to the European Convention, which will be signed in October this year, grants highest courts and tribunals the power to seek an advisory opinion, which Strasbourg will give without prior CJEU involvement. Do we really expect national courts to, first, refer a matter to Luxembourg and, second, to ask Strasbourg for advise – or vice versa? That may happen occasionally in validity disputes whenever the legality of EU legislation is at stake. But in cases of consistent interpretation, that will not happen often.
Forty years ago, the European Court of Justice was confronted with a situation where Member States considered the conclusion of an international treaty concerning a subject on which an EU regulation existed. The position of the court in Luxembourg was firm and courageous: Member States may not ‘assume obligations which might affect [EU] rules or alter their scope.’ The rationale underlying this position remains persuasive: to prevent situations in which national courts have to choose between Union law and international obligations by entrusting EU institutions with international representation. As a result, the Commission speaks within the WTO and in open skies negotiations – also in situations where Union law is implemented by national authorities.
The rules of attribution in the Draft Accession Agreement follow a different logic. They designate Member States as the primary respondent when they are implementing Union law. This solution may protect the CJEU’s interpretative autonomy on the division of competences between the European Union and its Member States, but it challenges the uniform interpretation and effective application of Union law. One may counter that the EU legal order is mature enough to sustain such contestation and this may very well be the case. But the challenge remains and should be articulated.
What will judges in Luxembourg decide in Opinion 2/13? A clear-cut response would be to instruct re-negotiations with a view to bring rules of attribution and the trigger for the co-respondent mechanism more in line with established case law. This need not cause much delay, since negotiations could be concluded swiftly, if Luxembourg presented a viable way ahead (the entry into force of the final text will require ratification by the EU and all 47 members of the Council of Europe anyway, which might take years). Alternatively, the Court of Justice may opt for technical improvements, such as effective Commission scrutiny of all complaints together with a quasi-automatic trigger of the co-respondent mechanism instead of a cumbersome Council decision under Article 218.6 TFEU. Irrespective of the outcome, the objective is evident: to support oversight of EU law by the ECtHR, while preventing that Strasbourg intervenes instead of Luxembourg.