20 December 2017

Taking the EU-Turkey Deal to Court?

The EU-Turkey Statement of 18 March 2016, also known as “the deal” or EU-Turkey deal, has been one of the most controversial policy steps taken by the EU in recent years. As such, it has also been the subject of a debate in the Verfassungsblog between two eminent scholars: Prof. James Hathaway and Prof. Kay Hailbronner. In the aftermath of the deal, intergovernmental organizations as well as human rights NGOs have issued reports on why Turkey is not a “safe third country”, and why the deal should not serve as a blueprint in negotiating similar deals with other third countries. A recently published study in the VU Migration Law Series clearly demonstrates the gravity of the situation of those who have been returned from Greece to Turkey, especially those of non-Syrian origin. Arbitrarily detained with no access to a lawyer or information, and with no possibility to apply for international protection, many face a serious risk of deportation to their countries of origin.

The deal was made public via a press release of 18 March 2016 under the title “EU-Turkey Statement” (Statement). Some of the most important commitments agreed upon were the return of all new irregular migrants from Greece to Turkey as of 20 March 2016 and the implementation of the “one for one” scheme under which for every Syrian readmitted to Turkey, one Syrian from Turkey would be relocated to the EU. Some of the other points mentioned in the Statement were upgrading the Customs Union (CU), the acceleration of the implementation of the visa liberalization Roadmap as well as Turkey’s accession process. As part of the deal, the EU also agreed to mobilize an additional €3 billion for the EU Facility for Refugees in Turkey by the end of 2018.

The deal raises many interesting legal and political issues. What fuelled further interest in it more than a year after its conclusion, were the orders of the General Court (GC or Court) delivered in proceedings brought by Pakistani and Afghan asylum seekers who were affected by it and sought its annulment (the NFNG, and NM cases). The Court chose to sidestep the difficult legal questions raised by the deal by dismissing these cases, ruling it had no jurisdiction to review the deal on the ground that the Statement was not an act of Union institutions, but that of Member States. Before discussing whether that conclusion was correct, it is worth casting a brief look at one of the orders (since they are almost identical).

The Order of the General Court in NF

The first striking element in the order of the Court is the way in which it sketches over the background of EU-Turkey cooperation in the area of migration. The Court’s account of events leading to the deal starts with a meeting between the EU leaders and Turkey of 15th of October 2015, skipping the single most important document on which cooperation in this area is built, i.e. the EU-Turkey Readmission Agreement (RA). This is a significant omission, as the Readmission Agreement is instrumental in deciding who had competence to act in this area.

The second surprising element in the order is the reaction of the institutions when the Court requested them to submit relevant documents to identify who concluded the deal, and whether the meeting of 18 March 2016 led to a written agreement. The European Council, the Council of the EU (Council) as well as the Commission all denied any involvement in the process leading to the deal, while there was plenty of evidence to the contrary in the media, and claimed that the Statement was “not intended to produce legally binding effects nor constitute an agreement or a treaty” (NF, para. 27). It was merely “a political arrangement” (NF, para. 29).

The third element that is worth pointing out is the fact that the Court merely paid lip service to the principle that “substance comes over form” in judicial review (NF, para. 42); it did not go into analysing the substance of the Statement. Eventually, it decided the case entirely on form. What was decisive for the Court were the invitations sent to various parties as well as the “Working Programme of the Protocol service”, which enabled it to conclude, “notwithstanding the regrettably ambiguous terms of the EU-Turkey statement” (NF, para. 66), that Member States acted in their capacity as Heads of State and Government on 18 March 2016. The fact that the President of the European Council and the President of the Commission were also present in the meeting did not change that conclusion, as they were not officially invited (NF, para. 67). Since the Statement was not an act of an EU institution, the Court had no jurisdiction to review it.

Lastly, the Court added, for the sake of completeness, that even if the deal had in fact been an agreement, the Court would not have had jurisdiction to rule on its lawfulness, as that agreement would have been concluded between the Heads of State and Government of the Member States of the Union and the Turkish Prime Minister (NF, paras. 72-73). But is that approach correct? Are Member States competent to conclude an agreement on their own on an issue (readmission) that has already been covered in an agreement recently concluded between the EU and Turkey? Are they competent to act on their own in such an area?

These questions require a brief discussion of the division of competences between the EU and its Member States, since the EU is a legal order based on the principles of the rule of law (Art. 2 TEU) and conferred powers (Art. 5 TEU). Hence, it is important to establish who had the competence to act in this specific case.

Who had competence to conclude the deal and what was the appropriate procedure?

To identify the appropriate procedure that had to be followed in concluding the so-called deal, one has to look at “the content and aim” of the Statement (Venezuelan fishing rights, para. 74). One can deduce from reading the Statement that its primary aim is to materialize the return of “all irregular migrants” to Turkey as of 20 March 2016. This means that the crux of the deal concerns the area of freedom, security and justice (Art. 4(2)(j) TFEU), which is an area of shared competence between the EU and its Member States (while many of the other points such as visa liberalization and upgrading the CU are matters of exclusive EU competence). The applicable provision is Art. 79 TFEU to which the ordinary legislative procedure applies (Art. 79(2) TFEU). According to Art. 218(6)(a)(v) TFEU, agreements covering fields to which the ordinary legislative procedure applies are to be concluded by the Council after obtaining the consent of the European Parliament. As far as the conclusion of RAs is concerned, the Union has expressly conferred powers to conclude such agreements under Art. 79(3) TFEU.

In the areas of shared competence, Member States are allowed to exercise their competence to the extent that the Union has not exercised its competence or to the extent it has ceased to exercise it (Art. 2(2) TFEU). As far as the readmission of third country nationals (TCNs) by Turkey is concerned, this is clearly and precisely what the EU-Turkey RA covers. In other words, the Union has exercised its competence in this specific area, pre-empting Member States’ competence to conclude an agreement with Turkey on that topic. It is worth noting that even if Art. 79(3) TFEU did not exist and the Union had no express external competence thereunder, by virtue of the existing EU-Turkey RA, the Union would still have an implied exclusive competence under the ERTA doctrine, which is now codified under Art. 3(2) TFEU, as the commitments undertaken on the return of irregular migrants under the deal “may affect common rules or alter their scope”. For an example, see the definition of the concept of a “safe third country” in Article 38(1)(e) of the Directive on common procedures to grant and withdraw international protection. For a concrete example of a change undertaken so as to implement one of the commitments under the deal (the “one for one” commitment), see Decision 2016/1754, which amended the second Relocation Decision adopted for the benefit of Greece and Italy so as to allow the relocation of Syrian nationals present in Turkey under the existing scheme.

It is also worth emphasizing that in the areas in which the Union has already exercised its competence, Member States are not only pre-empted from concluding international agreements as the wording of Art. 3(2) TFEU lays down, but they are pre-empted from taking any actions that might lead to the adoption of acts with legal effects. For instance, in the International Maritime Organization (IMO) case, Greece failed to fulfil its obligations under the Treaties by submitting a mere proposal to the Maritime Safety Committee of IMO, as that could lead to legally binding rules over time. This case provides a good analogy for the EU-Turkey deal. For those thinking that the Statement is not problematic because it is not legally binding, it is easy to point to the laws passed by Greece within a few weeks of its release to provide for the effective implementation of the deal. To sum up, irrespective of the legal nature and immediate effects of the Statement, I believe Member States had no competence/power to act on their own concerning the return of irregular migrants to Turkey.

What next?

The three orders have been appealed to the Court of Justice of the EU (CJEU). The strict admissibility requirements laid down in the Plaumann ruling place an important constraint on the CJEU. If the case had been brought by the European Parliament, which was entirely sidestepped in the process leading to the deal, instead of dealing with issues of admissibility, the CJEU would be looking at the substance of the matter. This case illustrates how the checks and balances built into the system can be completely bypassed when the EU institutions collude with Member States to act outside the Treaty framework.

In this particular case, one sees the value of the system being one of “dual vigilance”, in that it is not only the Commission (under Art. 258 TFEU) or the other institutions as privileged applicants that have access to the Court, but also individuals. Even though the term is usually employed in the context of the preliminary reference procedure, we see how important the role of individuals can also be in the context of the annulment procedure (Art. 263 TFEU), especially when locking out a preliminary reference from a national court does not appear to be a realistic option. Thus, it is important that the CJEU use this opportunity to set the record straight by establishing who had the competence to conclude the EU-Turkey deal.

Why is it important that the Court of Justice answer the question on competences? Firstly, because the EU is a legal order based on the rule of law and conferred powers, which Member States seem to forget when they think there is a “crisis”. The CJEU should send a clear signal that rules are there to be respected at all times. Secondly, because it is the duty of the CJEU to ensure that Member States (Art. 4(3) TFEU) as well as Union institutions act in good faith and within the powers conferred on them by the Treaties (Art. 19(3) TFEU). Thirdly, protection of fundamental rights and the rule of law form part of the very foundations of the Union legal order (Kadi I, paras. 303-304). Dismissing the important questions raised by the deal, as the GC did, will undermine those very foundations and will cast doubt on the existence of an effective system of judicial protection in the EU legal order.

This blog post is based on a T.M.C. Asser Policy Brief and a longer academic article.