Some thoughts on the ECJ hearing on the Draft EU-ECHR accession agreement (Part 1 of 2)

I was in Luxembourg 5-6 May 2014, attending the hearing at the European Court of Justice concerning the Draft Agreement for the Accession of the European Union to the European Convention on Human Rights. In these two posts (click here for part 2) I summarize the main arguments presented at the hearing, and provide some initial analysis and thoughts. Finally, I would not be a lawyer if I did not point out that some caveats apply: I am writing based on my own notes and recollection, and thus there might be inaccuracies, omissions and misattributions. If you come across any, feel free to point them out in the comments section below. I would also greatly appreciate any other comments or thoughts you might have in relation to this case.

Monday May 5th was the opening day for the case concerning the Draft Agreement for the Accession of the EU to the European Convention of Human Rights (hereinafter: the DAA). The case is brought under the procedure provided for in TFEU article 218(11) by the Commission, which is asking the European Court of Justice (ECJ) the question of whether the DAA is compatible with the constituent treaties of the European Union. Much could be (and has been) written about this question, this case, and the spectacle that is an ECJ hearing before the full court. In these couple of posts I will, however, focus on the submissions of the parties. I will give you what I perceived as the highlights of the hearing, and provide some initial commentary.

Let’s start off with the most important fact of the first day: all three EU institutions and all the 28 EU member states submitted that the DAA is indeed compatible with the EU treaties. This consensus was not unexpected, given that all the EU member states have unanimously already agreed to the text of the DAA. Due to this consensus, the proceedings before the ECJ were of a quite peculiar nature: Thirty-one agents were trying to convince the ECJ that the DAA does not encroach upon the ECJ’s own position and prerogatives under TEU article 19, its monopoly on the interpretation of Union law under TFEU article 344, or the broader concept of the autonomy of the Union’s legal order. (See e.g. the EEA opinions (Opinion 1/91 &  Opinion 1/92), and, more recently, Opinion 1/09 Creation of a unified patent litigation system [2011]).

Possibly as a result of this lack of contradiction, the ECJ had communicated a list of questions to the parties that were to serve as the basis for the oral hearing. Although this list questions does not seem to have been made public, I will touch upon most, if not all of the topics they concern, in these two blog posts. It should also already at this stage be noted that in answering these questions it was possible to observe several sever splits and minor disagreements between the EU institutions and/or between the EU member states.

The relevance of the internal EU rules
There was some disagreement between the member states as to the relevance of the so-called “internal rules” to the present proceedings. These internal rules are the rules that will implement the DAA in Union law. Their content is presently undecided (a point which I will come back to). But, they will probably include detailed rules on the practical application of the co-respondent mechanism, representation at the European Court of Human Rights (ECtHR), apportionment of responsibility, obligations of the EU member states to inform the Union about relevant ECtHR cases, and so on.

It was rather surprising – at least to me – that no draft of such rules were available at present. During the negotiations of the DAA one got the impression that a draft of these rules would be available before the ECJ gave its opinion.

The relevance of this lack of draft internal rules to the present proceedings proved contentious. Some parties were of the opinion that the ECJ should take into account the possibility that such internal rules could mitigate the problems that the ECJ may have with regard to the DAA/accession. Other parties (and probably the majority), including the UK, Sweden, the Netherlands and Spain, submitted that the internal rules were entirely irrelevant. According to them, taking into account possible future rules that have not yet been agreed would go against the division of powers between the Union institutions. Several of those member states also argued that any assessment of the (possible or likely) internal rules would fall outside the scope of the ECJ’s jurisdiction under TFEU article 218(11), since that would entail going beyond checking the compatibility of the DAA with the Union’s constituent treaties. Pursuing this line of argument further, some of these member states also submitted that the internal rules would also be irrelevant for the ECJ’s assessment, because they can not overrule the provisions of the DAA, which is an international treaty between 48 parties (the 47 state parties to the European Convention on Human Rights (ECHR) and the Union).

While i (at least partly) subscribe to the arguments that the internal rules are in principle irrelevant for the present proceedings, I do not think that the ECJ can completely ignore the opportunities such rules may provide. Here I think the EU Council took the most sensible approach: It submitted that the ECJ may find the treaty compatible given that certain provisions, or safeguards, are adopted as part of the internal rules. In my opinion this must be correct, and I do not think that the ECJ need to even mention the concept of “internal rules” in its opinion. Rather, since the ECJ is able to declare the DAA both compatible or incompatible with the constituent treaties of the Union, it should also be able to find a middle ground: upholding the DAA with some conditions or clarifications. These conditions could then be translated into internal rules at a later stage, or be sorted out through other appropriate means.

The “co-respondent” and “prior involvement” mechanisms
A particular feature of the DAA is the co-respondent mechanism, which gives the Union a right to join the proceedings against an EU member state before the ECtHR in a case where the compatibility of EU law is called into question (DAA article 3). When this mechanism is triggered, the ECJ will also be given an opportunity to assess the comparability of the relevant provision of Union law, if it has not already made such an assessment at an earlier stage (DAA art. 3(6)). This is called the prior involvement mechanism. Although there was some debate concerning the prior involvement mechanism itself – for example whether prior involvement would be permissible if the ECJ had already given a preliminary opinion to a domestic court in the same case – most of the debate focused on the triggering and practical application of the co-respondent mechanism.

One question that came up was whether the member states would be under an obligation to inform the Union’s organs when they received cases communicated to them from the ECtHR where the compatibility of Union law with the ECHR could be called into question. Such information is crucial if the EU is to make use of the co-respondent mechanism in practice. The parties to the case split into three groups on this question. Some argued that the EU member states would be obliged to inform the Union’s organs of relevant ECtHR cases under the duty of sincere cooperation (TEU article 4(3)). Others argued that an obligation to inform the Union’s organs would have to form part of the forthcoming internal rules. Yet others submitted that there would be no obligation on the member states to inform the Union’s organs, and that it would be up to a political decision whether to include such an obligation in the internal rules – thus not ruling out that the Union would itself be tasked with unearthing those (relatively few) cases in the ECtHR’s overfilled docket that raise issues of the compatibility of Union law with the ECHR.

I have much affinity for the argument building upon the duty of sincere cooperation. And I would be surprised if an obligation to inform the Union’s organs about relevant communicated cases is not included in the forthcoming internal rules. The member states seem to be in a much better position to identify such issues as the cases are communicated to them from the ECtHR compared with the Union, which would have to primarily rely on the public documents available at the ECtHR’s website. Moreover, it would most likely be in an EU member state’s interest to inform the Union, since bringing the Union in as a co-respondent could possible relieve the member state of its responsibility under the ECHR.

When the Union does obtain information about a relevant case at the ECtHR, a further question arises: Is the Union is obliged to join the ECtHR proceedings as co-respondent? With regard to this question the EU member states were split. The Commission nevertheless argued that the EU had an obligation to join the proceedings in every relevant case. This view was also shared by the European Parliament. This view is nevertheless difficult to square with the text of DAA article 3(2), which provides that “the European Union may become a co-respondent” (emphasis added) in cases where the incompatibility of Union law is alleged. This is somewhat reinforced by the draft declaration that the Union should make upon signing the DAA, and the Draft model of Memorandum of Understanding between the EU and its member states. These two instruments were added to the DAA package at the very end of the negotiations, during the final negotiating meeting. Both instruments aim to assure the Union will seek to join as co-respondent in all such cases. Again, though, one should expect that an obligation to join all relevant cases as co-respondent will be given a firm and clear legal basis in the forthcoming internal EU rules.

If the EU decides to join as co-respondent, it would have to submit an application to the ECtHR (unless it is directly invited by the ECtHR itself, see DAA art. 3(5)). This application is to be decided by the ECtHR. Thus the question arises as to whether it is compatible with the autonomy of the Union’s legal order, and the ECJ’s interpretative autonomy under TFEU art. 344, for the ECtHR to decide whether or not a case before it calls into question the compatibility of EU law with the Convention. This is because that in doing so, the ECtHR may run the risk of having to interpret EU law to reach its decision – or so the argument goes.

However – as pointed out almost unanimously by the parties to the case, albeit in different variations – the ECtHR is not really put in such a position. According to DAA art. 3(5) cf (2) the ECtHR should only assess “whether, in the light of the reasons given by the [Union], it is plausible” that the requirement that “an alleged violation … appears [to] call into question” the compatibility of a provision of Union law with the ECHR is fulfilled (emphasis added). As pointed out by e.g. the UK, this is an exceptionally low standard, tailored to safeguard the Union’s legal autonomy. The ECtHR will only assess the reasons given by the Union, not Union law itself. Moreover, the ECtHR will only have to find that the reasons provided to it makes it plausible that there is an apparent conflict between EU law and the ECHR. It therefore seems that the Union’s legal autonomy will not be threatened by the ECtHR’s power to decide on applications to become co-respondent. After all, it seems rather appropriate that the ECtHR gets to have final word, however rarely it may use it, as for who is allowed as (co-)respondents before it in a case.

A final contentious issue under the co-respondent mechanism was the apportionment of responsibility between the respondent(s) and the Union as co-respondent. The Commission argued that according to DAA article 3(7) the ECtHR would be obliged to find the respondent(s) and the co-respondent Union jointly responsible, unless the Union and the respondent member state(s) agree to apportion the responsibility between them. This solution will, according to the Commission and those supporting its line of argument, safeguard the autonomy of Union law by making sure that the ECtHR would never have to decide whether a violation of the ECHR is the result of a freely decided and autonomous act of an EU member state, or an act ordered by binding obligations of EU law. This line of argument was joined by many parties, often with reference to paragraph 62 of the Explanatory Report to the DAA.

I am, however, not entirely convinced by these arguments. Already today the ECtHR has shown that it is willing to delineate between those acts that an EU member state is obliged to perform under Union law (think Bosphorus Airways v. Ireland [GC], no. 45036/98 (2005)) and those cases where the EU member state could have avoided violating the ECHR if it had used its discretion in implementing Union law differently (think e.g. M.S.S. v. Belgium and Greece [GC], no. 30696/09 (2011)). A clearly worded provision to the contrary would therefore seem necessary in order to for the ECtHR to be obliged to deviate from its present practice. The text of DAA article 3(7) does not in my mind meet this threshold. At best, it is ambiguous: if a violation is found in these cases, DAA article 3(7) provides that “the respondent and the co-respondent shall be jointly responsible for that violation, unless the Court, on the basis of the reasons given by the respondent and the co-respondent, and having sought the views of the applicant, decides that only one of them be held responsible” (emphasis added). Nothing here seems to require an agreement between the Union and the other respondents.

The key wording, which I have highlighted in the quote above, would seem equally fitting in a case where the Union and one or more member states disagree as to whether they are jointly liable. For example if an EU member state argues that it should not be held responsible because its conduct was ordered by the Union – which might entail only the responsibility of the latter. Perhaps this cannot be the case given how the Bosphorus judgment of the ECtHR is formulated. But the opposite situation, where the Union argues that only the member state should be responsible, while the member state disagrees and argues for joint responsibility, could possibly occur. And that situation would also seem to fit the key wording underlined in the previous paragraph. Moreover, while the Explanatory Report paragraph 62 might be more explicit, supporting the Commission’s interpretation, I submit that it will not necessarily control the interpretation of DAA article 3(7). The Explanatory Report is only a supplementary mean of interpretation, and should therefore be given very little weight. And, if I remember correctly, an earlier draft clearly providing for joint responsibility in lieu of explicit agreement to the contrary was rejected in the final stages of negotiations. Paragraph 62 of the Explanatory Report thus seems to be just the final relic that remains of that rejected provision. Finally, paragraph 62 of the Explanatory Report may just as well be read as a provision of non-binding advice to the ECtHR. Other than reiterating the language of DAA article 3(7) it does little other than warn against the risks of an overzealous ECtHR to the autonomy of the EU legal order.

The ECJ’s (lacking) jurisdiction over the Common Foreign and Security Policy, and ECHR article 13
The EU’s Common Foreign and Security Policy (CFSP) was part of the reason why a new round of negotiations of the DAA had to be undertaken in 2012/2013. According to leaks from the EU Council, particularly France was worried about giving the ECtHR jurisdiction in relation to acts adopted by the Union under the CFSP. An amendment to the DAA that would exclude CFSP acts from the ECtHR’s jurisdiction was however rejected by the non-EU state parties to the ECHR who insisted that the EU should join in full, with no parts of Union law to be excluded. Only this would properly respect the principle that the Union should accede on an equal footing with the state parties to the ECHR. Finally, a deal was brokered whereby there would be no exception to the ECtHR’s jurisdiction in relation to CFSP acts, in exchange for some adjustments to the language in DAA article 1(4) – a topic which I will come back to.

Including the CFSP within the jurisdiction of the ECtHR could be problematic because, according to the traditional view, the ECJ has a very limited jurisdiction over CFSP acts. As laid out by France in its submission, and joined by the UK, several other member states, the EU Council, and the European parliament, the primary rule with regard to the CFSP is that the ECJ does not have jurisdiction (TFEU article 275(1)). The only exceptions are so-called “restrictive measures” under TFEU article 275(2), normally interpreted as referring to sanctions, and the savings clause in TEU article 40. Support for this view may also be found in Case C-354/04 P SEGI et. al. [2007], where an action for damages under what is now TFEU article 263 was, according to France, denied under the former third pillar. France also referred to Case C‑583/11 P Inuit Tapiriit Kanatami et. al. [2013], where the ECJ refused to interpret a changed wording of a jurisdictional clause in the Lisbon treaty in an expansive manner – instead reaffirming the Plaumann doctrine, despite the more inviting post-Lisbon language of the disputed provision.

Due to this limited jurisdiction, there is a risk that the ECJ would not be able to interpret the applicable Union law before the ECtHR is to rule on a case calling into question a CFSP measure with the ECHR. The “prior involvement” mechanism would in other words only result in an dismissal from the ECJ. Thus, the ECtHR would be on its own, possibly left to interpret Union law without guidance. France provided a hypothetical example of this, namely if in an EU military action there is collateral damage: could those affected sue the Union before the ECtHR? (The Behrami and Saramati cases before the ECtHR spring to mind). The French reply would be yes.

This can potentially threaten the autonomy of Union law, and the ECJ’s interpretative monopoly under TFEU article 344. Moreover, the lack of judicial protection at the EU level could in such cases also entail a breach of ECHR article 13, which obliges the Union to provide an effective remedy in cases where ECHR rights are violated. As to the latter, France and its allies in the hearing argued (with some variation) that sufficient protection would be afforded by the national courts and the ECJ’s (exceptional) jurisdiction over these matters. In this respect, France put emphasis on article 1(4) of the DAA which it seems to read as a rule attributing responsibility for acts under the CFSP to the EU member states, and not the Union.

The Commission suggested a very radical, and surprising solution to these difficulties. It submitted, in contrast to France and its supporters that the ECJ’s jurisdiction is not as narrow in CFSP cases as traditionally held. In this view it was joined, at least partially, by Germany. The Commission argued that the EU is a community of law. In such a community, the ECJ must have jurisdiction in relation to CFSP acts when the rights of individuals are affected. Germany explicitly referred to the Kadi I case as an example of previous expansive interpretation along these lines, as well as case C-91/05. The Commission also endorsed a very expansive interpretation of the term “restrictive measures” in TFEU article 275(2), so as to include most cases of human rights violations that may arise under the CFSP, as an alternative ground for ECJ jurisdiction in such cases. Germany added that with respect to treaties under the CFSP, the ECJ would have jurisdiction to assess their compatibility with fundamental rights under TEU article 37 cf. TFEU article 218. Furthermore the Commission pointed to the possibility of cases under TFEU article 268 for non-contractual liability when CFSP acts are in play. The Commission cited the SEGI case as affirming this, thus reading it in a different way than France, but exactly on what point they disagree was left unclear (at least to me, as I am not too familiar with the SEGI case).

From the above, we can see France is arguing that the Union would escape the ECtHR finding of breach of ECHR article 13 for two main reasons: (a) acts under the CFSP should be attributed to the member states under DAA article 1(4), and (b) domestic courts will deal with those cases in a way that provides appropriate remedies. This reading of DAA article 1(4), apparently seeing it as a rule on exclusive attribution, is not very convincing in my mind. Article 1(4) is generally formulated, and seems to do nothing more than affirm the first aspect of the ECtHR’s Bosphorus case, namely that EU member states are responsible when they implement Union law, even if they are afforded no discretion as to its implementation (Bosphorus, at paras 136-137 and 153). It does not exclude that the same conduct could be attributed also to the Union (dual attribution), or that the Union could be held responsible for a different but related conduct (giving raise to the same violation/damage). For example, the Union would be responsible for the act of giving an illegal order, by way of a binding legal act, while the member state would be held responsible for the act of physically implementing that act. In such cases, there are good arguments that the Union and the member state in question would be jointly liable (shared responsibility).

On the other hand, the arguments of the Commission and Germany seems equally ambitious, since they are in practice asking the ECJ to assert its jurisdiction over sensitive areas of policy that the drafters of the Lisbon treaty seem to have intended to keep out of the ECJ’s jurisdiction.

It will be interesting so see how the ECJ will handle these issues in its opinion. It could act boldly, and broaden its jurisdiction, but that seems unlikely. Even for the ECJ it would seem a too dramatic display of competence creep/dynamic interpretation. If it lacks the necessary boldness it might, in a worst case scenario, opine that accession to the ECHR would require a treaty change so as to give the ECJ jurisdiction over the CFSP with regard to fundamental rights. This conclusion would of course be disappointing, as it would set the accession procedure back by years – given that it would be at all possible to pass such a treaty amendment. That the ECJ will choose follow the French interpretation of DAA art. 1(4) also seems unlikely, since it seems to lack support in the text. Also, DAA art 1(4) would in practice be applied by the ECtHR. It seems particularly surprising if the ECtHR would subscribe to France’s rather peculiar interpretation of that provision. At the very least, the ECJ will probably not entrust the ECtHR to interpret it in such a peculiar way.

Perhaps a more workable solution would be a middle ground approach whereby it is accepted that both the Union and an implementing EU member state could be jointly responsible (thus discarding the French interpretation of DAA art 1(4)), but at the same time betting on the fact that the ECtHR will accept that national courts in cooperation with the ECJ (in the few cases where it has jurisdiction) will be sufficient under ECHR article 13 (thus discarding the Commission’s and Germany’s calls for the ECJ to assert a sweeping jurisdiction over the CFSP). This would, however, leave the risk of threatening the EU’s legal autonomy and the ECJ’s interpretative monopoly under TFEU article 344. As to the latter point, the ECJ might find that TFEU article 344 is irrelevant, as the ECJ indeed does not have jurisdiction in the cases where it cannot make good use of the “prior involvement” mechanism. As to the autonomy of EU law, however, I do not have any good suggestions. But, on the other hand, it seems fairly unclear what the content of this principle is. Previous case-law of the ECJ on the matter is rather inconclusive (if not erratic). Sometimes it seems like the “autonomy argument” is a tool for the court in relation to agreements it does not like. Therefore I predict it will not pose a great challenge here, as the ECJ itself has showed strong interest in the Union’s accession to the ECHR. (For a bit more detail on this, see part 2 of my analysis of this hearing).

Finally, that some acts under the CFSP could be attributed to the Union and lead to violations of ECHR article 13 due to the lack of remedies at the Union level is in itself not something that should block accession. The ECHR does not prohibit parties from joining it if they have not fulfilled all the rights embodied in it. In fact, I would go as far as to claim that all ECHR state parties most likely had laws and practices that were in obvious violation of the convention at the time they acceded to it. The Union could fix its problem through a treaty revision at a later stage, given that the ECtHR actually finds that the judicial protection is insufficient in concrete cases before.

The omissions
Some aspects of the DAA were not discussed, or discussed less than I expected. Two such omissions stick out for me: the future of the Bosphorus presumption, and the “reversed co-respondent cases”.

As for the first point, it is not terribly surprising that the Bosphorus presumption was not discussed. It was probably not on the ECJ’s question sheet, and it is more relevant for the ECtHR than the ECJ. But it could have been discussed in relation to the CFSP, as Sweden half-heartedly attempted to do towards the very end of the hearing – hinting that the Bosphorus presumption would survive accession, at least when it comes to cases under the CFSP.

The other point, “reversed co-respondent cases” is my label for those cases where the Union is sued directly before the ECtHR. In other words where the EU is the respondent, and the member states may (possibly) become co-respondents. The condition here is however that the case concerns incompatibilities between the constituent treaties of the Union and the ECHR, see DAA art. 3(3). These kind of cases were not discussed at all, other then passing mentions, but I think the deserve attention before the ECJ.

Questions to the parties before the second day of hearings
The ECJ did not close the hearing before 7 pm, but nevertheless gave the parties extensive homework before tomorrow. Both the judge-rapporteur, General Advocate Kokott, and President Skouris had questions they wanted answers to.

The judge-rapporteur asked a long-winded question concerning the relationship between the ECJ and the ECtHR and the effects of this reciprocal relationship on the Union’s legal autonomy. In particular, he asked if one could envisage a situation where the Union could “react” to Strasbourg case-law (i.e. refuse to comply) in similar fashion as the apex courts in (some) ECHR state parties. In my view, the answer here would probably be up to the ECJ itself. The Union is to be put on an equal footing with state parties to the ECHR, as emphasized in the DAA, and thus the ECJ should be in the same position as any other domestic court.

Advocate General Kokott asked a brief question, after noting that the DAA subjects the Union to ECtHR jurisdiction with regard to human rights: Is this compatible with the autonomy of Union law. This seems like a fairly vague, but extensive question, probably to try to lift focus up from the nitty gritty details to more cross-cutting general principles.

The President Skouris asked two questions. First, whether acts termed administrative in the EU context could be re-characterized by the ECtHR, for example in competition cases, thus triggering ne bis in idem issues. In my humble opinion the answer here is obvious: yes, and it would not affect neither the autonomy of Union law or interpretative monopoly of the ECJ. Looking beyond national labels is one of the core principles of how the ECtHR operates.

The President’s second question was what the position of the parties was to ECHR protocol 16. My guess is that since that protocol is optional, and the DAA does not envisage the EU joining that protocol in the course of accession, protocol 16 would not be relevant for the present proceedings. Instead the question of ECHR protocol 16 and its compatibility with the co-respondent mechanism et cetera should be the subject of its own opinion when and if the Union decides to join it at a later stage.

The second and final part of my summary and analysis of the EU-ECHR accession hearing is available here. This post was slightly edited, corrected, supplied with additional links, and (somewhat) proofread on 9 May 2014.