Abstract: In its judgment in the Swiss MoU case, the Grand Chamber of the European Court of Justice (ECJ) annuls a European Commission decision on the signing of an addendum to a Memorandum of Understanding with Switzerland on the subject of Switzerland’s financial contributions to ensure that country’s continued access to the internal market after Croatia’s accession in 2011. In doing so, the ECJ makes a meaningful contribution to clarifying the law on the issue of the division of powers between the EU institutions with regards to the conclusion of non-binding agreements by the European Union – an issue on which the EU Treaties are silent. Unfortunately, the picture to which Swiss MoU adds a few brush strokes is one in which the power to conclude non-binding agreements is divided only between the Commission and the Council. The European Parliament, by contrast, remains invisible. This Insight argues that a more comprehensive reading of the EU Treaty is called for; one which takes into account not only the policy-making powers of the Council and the power of external representation of the Commission, but also the power of the European Parliament to exercise political control and legislative functions. Only such a reading would do justice to the dual source of democratic legitimacy of the EU: the individual EU citizens as represented in the Parliament, and the EU citizens as represented by their governments in the Council.
Keywords: EU external action – non-binding international agreements – division of competences – institutional balance – conferral – accountability.
On 28 July 2016, the Grand Chamber of the European Court of Justice (ECJ) rendered judgment in the Swiss Memorandum of Understanding (“Swiss MoU”) case. The case offered the Court an opportunity to further clarify the division of powers between the political institutions of the EU in the area of external action. In particular, the Court provided some clarification on the procedural legal basis on which the EU can conclude so-called ‘non-binding’ or ‘political’ agreements with third countries – an issue for which the Lisbon Treaty did not make express provision.
The immediate cause of the dispute in Swiss MoU was the adoption, on 3 September 2013, by the Commission, of a decision authorising the Commission Vice-President and the Commissioner for Regional Policy to sign an MoU with Switzerland (the “2013 MoU”). The MoU served to establish a legal framework within which Switzerland would be able to conclude bilateral agreements with Croatia. These bilateral agreements would allow Switzerland to finance projects within Croatia, as part of Switzerland’s financial contributions to ensure continued access to the EU internal market after the accession of Croatia to the EU in 2011.
The 2013 MoU was not the first of its kind. Similar MoU’s had been concluded on the occasion of earlier accession rounds. The 2013 MoU, however, was the first MoU to be adopted with Switzerland after the coming into force of the Lisbon Treaty, in 2009. In contrast to the earlier MoU’s, the 2013 MoU was concluded and signed by the Commission independently, without Council participation. The Commission felt confident to proceed this way, as in 2012 the Council and the representatives of the Member States had adopted conclusions in which they had invited the Commission, in close cooperation with the Presidency of the Council, to “engage in the necessary discussions” with Switzerland to come to an agreement on that country’s financial contribution to compensate for the enlargement of the EU internal market after Croatia’s accession. The Council conclusions required the Commission to consult the relevant Council working group as to the progress of the discussions.
As mentioned, the Lisbon Treaty does not provide for the conclusion of non-binding agreements with third countries. This raised the question of the appropriate procedural legal basis upon which such decisions authorising the signing of such agreements ought to be adopted. The Commission considered itself empowered to adopt the decision at issue on the basis of its powers to ensure the EU’s “external representation”, and its power to fulfil “executive” as well as “management” functions. All of these powers are given textual expression in Art. 17, para. 1, of the TEU. The Council, by contrast, understood the decision to conclude the MoU as an exercise of its power to carry out “policy-making functions”, as provided for in Art. 16, para. 1, TEU.
In short, the constitutional questions at issue in Swiss MoU are the following: how is one to read the scope of the Commission’s and Council’s powers with regard to the conclusion of non-binding agreements with third countries? Where does “external representation” end and “policy-making” start in this increasingly important area of EU external action?
II. Advocate General Sharpston’s opinion
In a relatively elaborate opinion, AG Sharpston recommended the Court to annul the contested decision, while maintaining its effects until the adoption of a new decision.
Her opinion consisted essentially of four parts. In a first part, the AG examined the admissibility of the annulment action ex proprio motu. She wondered, in particular, whether a non-binding MoU as the one at issue should be considered an act of the institutions that “intended to produce legal effects vis-à-vis third parties” in the meaning of Art. 263 TFEU. The question of what should be understood by the term “legal effects” is not self-evident, as the discussion in the IOV case on the question of whether recommendations issued by the International Organisation for Wine and Vine should be considered as “having legal effect” in the meaning of Art. 218, para. 9, TFEU has illustrated. Drawing, in part, on the discussion in IOV, the AG argued in favour of admissibility, considering that the effect of the contested Commission decision on the powers of the Council can be understood as a “legal effect” vis-à-vis a “third party” in the meaning of Art. 263 TFEU.
In a second part, the AG touched on the issue of the division of powers between the EU and the Member States. Neither the Commission nor the Council, nor, for that matter, any of the intervening Member States, had raised this issue. The AG nonetheless felt compelled to address it, as the question of the division of powers between the institutions only arises in the presence of a substantive EU competence. Put differently, if the EU, taken as a whole, lacks power, a fortiori the individual institutions are not empowered to act. However, as the Council had not formally made a plea based on a lack of substantive competence for the EU to act, the AG considered the Council’s arguments on that issue to be inoperative.
The third part of the AG’s opinion addressed the core charge of the Council, i.e. its argument that, in adopting the decision, the Commission had encroached on the Council’s power to carry out “policy-making functions”. The AG answered this question in the positive. She considered that
“policy-making includes the decision that an objective for which the Union is competent can be pursued by obtaining a commitment (whether or not binding) from a third State to pay a financial contribution to a new Member State pursuant to a future bilateral agreement between those two parties (assuming no such decision has been taken earlier) and thus by participating in external action, in the form of negotiations and possibly the subsequent conclusion of an instrument to obtain that commitment.
When the Council has exercised that prerogative by authorising negotiations, it is then for the Commission to represent the Union in negotiations, in accordance with the Council’s authorisation and the Union’s policies and interests. However, that initial Council decision does not extinguish the Council’s power under Art. 16(1) TEU to decide on whether or not the Union should become a party to the instrument resulting from those negotiations and sign it.
It is for the Council to verify the content of the agreement, the form of external action used, whether any relevant constraints have been respected and the continuing need for the Union to become a party to that agreement”.
On this basis, the AG considered that, by authorising the signing of the MoU to take place, without obtaining prior Council approval, the Commission had encroached upon the Council’s policy-making authority. In doing so, the Commission had denied the Council the opportunity to assess whether the end result of the negotiations fitted within the Council’s broader policy on the topic of Switzerland’s financial contributions to ensure internal market access.
In a fourth part, the AG addressed the Council’s plea on the basis of the principle of sincere cooperation. The Council had essentially argued that, by refusing to obtain Council approval to the adoption of the contested decision, the Commission had violated the principle of sincere cooperation provided for in Art. 13, para. 2, TEU. Briefly put, while emphasising the importance of information sharing, the AG nonetheless considered the content of the Council’s plea on the basis of Art. 13, para. 2, TEU to correspond with the content of its earlier plea, alleging an encroachment on its policy-making powers.
As the AG considered the Council’s plea alleging a violation of Art. 16, para. 1, TEU well-founded, she proposed the Court to annul the contested decision. In order to avoid negative repercussions on the EU’s relations with Switzerland, she advised the Court to maintain the contested decision’s effects until the adoption of a new decision.
III. The Grand Chamber’s judgment
The Grand Chamber proceeded to annul the contested decision, while maintaining its effects until the coming into force of a new decision, which ought to be adopted “within a reasonable time.”
Despite the AG’s elaborate treatment of the question, the ECJ did not dwell on the question of the action’s admissibility. Instead, it immediately turned to the substance of the case, i.e. the Council’s pleas alleging an encroachment upon its powers to make policy and a violation of the principle of sincere cooperation. In response to the first plea, the Court held that
“[t]he decision concerning the signing of an agreement with a third country covering an area for which the Union is competent — irrespective of whether or not that agreement is binding — requires an assessment to be made, in compliance with strategic guidelines laid down by the European Council and the principles and objectives of the Union’s external action laid down in Art. 21(1) and (2) TEU, of the Union’s interests in the context of its relations with the third country concerned, and the divergent interests arising in those relations to be reconciled”.
This need to make an “assessment of interests”, the Court held, implies that the issue at hand could not be understood merely as a matter of external representation. Rather, the decision had to be understood as “one of the measures by which the Union’s policy is made and its external action planned for the purpose of the second sentence of Art. 16(1) and the third subparagraph of Art. 16(6) TEU”.
To the Commission’s argument that it had acted only within the boundaries set out by the Council in its 2012 conclusions, the Court responded that “the signature of a non-binding agreement entails the assessment by the Union of whether the agreement still reflects its interest, as defined by the Council in particular in the decision to open negotiations on the conclusion of the agreement”. In other words, the Council’s policy-making powers do not expire after the authorisation to negotiate an agreement; also at the phase of the conclusion, including the agreement’s signing, are policy-assessments involved – decisions which only the Council is to make.
As the Court decided to annul the Commission’s decision for reasons related to the division of powers between the EU institutions, it was not necessary to answer the Council’s second plea on the principle of sincere cooperation.
IV. Assessment: accountability, not only towards the Member States, but also towards the citizens of Europe
Non-binding agreements are concluded often to infuse flexibility in a polity’s foreign relations. Such flexibility is achieved, in many instances, by by-passing the legislature and thus by strengthening the executive. In many polities, the constitutionality of such practices is not contested. The lack of binding character of such agreements, it is typically held, ensures that they do not disturb the balance of power between the different branches of government.
The political importance of such non-binding agreements in contemporary international relations is, however, well-established. As authors such as Jan Klabbers have argued, moreover, such instruments of “soft law” tend to have rather “hard” legal effects once concluded, as courts more and more often rely on sources of non-binding, “soft” law in their reasoning. It is thus to be commended, and not entirely unsurprising in light of the ECJ’s long-standing tradition of purposive reasoning, that the ECJ did not deny the parties in Swiss MoU access to the court room merely because the agreement lacks “binding” force in the traditional, formalist understanding of the term. However, that the ECJ did not feel compelled to address the question of the admissibility of the action directly is, also in light of the AG’s elaborate treatment of the question, regrettable, as it confirms the ECJ’s reputation as a court that evades legal accountability.
On the substantive question of the division of labour between Council and Commission in the process of concluding non-binding agreements, the ECJ does not provide much justification for its decision to annul the Commission’s decision. In particular, a theory of how the Council’s policy-making power relates to the Commission’s power of external representation is still lacking. The only indication given by the Court is the reference to the reconciliation of “interests”. As the decision to authorise the signing of the 2013 MoU involved some form of interest assessment, the Council was empowered to adopt the decision, not the Commission.
Merely on the basis of this line of reasoning, it is not immediately clear what scope for independent action remains for the Commission. To acquire a fuller understanding of how the Court understands the relationship between the Council’s power to make policy and the Commission’s power to represent the EU externally, it is therefore interesting to zoom out slightly, and read the Court’s ruling in Swiss MoU alongside its judgment in the case of Australian Greenhouse Gas Emissions. In the latter case, the ECJ partially annulled negotiating directives adopted by the Council in the context of a treaty-making procedure. In the contested directives, the Council had empowered a committee to issue “detailed negotiating positions”, which the Commission was to defend in the treaty-negotiation process. The ECJ considered this part of the directives too detailed for the purpose of Arts 218, para. 4, TFEU and 17, para. 1, TEU and proceeded to partial annulment.
The broader picture emerging from a joint reading of Swiss MoU and Australian Greenhouse Gas Emissions is one in which the ECJ recognises the remarkably broad scope of the Council’s policy-making functions in treaty-making processes, be they of a formal or an informal nature. As soon as some form of “interest calculation” is present – and when is it not? – authority devolves to the Council. This implies that both the opening and the closing of treaty negotiations require a Council decision. During the treaty-negotiation process, however, the Commission enjoys a meaningful space for independent action, allowing it to anticipate and respond to positions and strategies employed by the other parties around the negotiating table. The Council is allowed to play a supervisory role by installing a special committee to which the Commission must occasionally report, but this supervisory role cannot go as far as to entail the issuing of binding detailed negotiating positions.
This arrangement is to be commended for the balance it strikes between protecting the Commission’s flexibility at the negotiating table on the one hand, and ensuring that the Commission remain accountable to the legislative branch – here the Council – on the other. The EU arrangement on this issue compares positively to arrangements in place in other federal-type polities, where, as mentioned, the executive typically is much less constrained by the legislature.
This observation, however, immediately points to the weakness of the picture that emerges from a joint reading of Swiss MoU and Australian Greenhouse Gas Emissions: ensuring political accountability also in the increasingly important area of “non-binding” political agreements requires not only accountability vis-à-vis the Member States, but also vis-à-vis the EU citizenry, as represented in the European Parliament. A more comprehensive and ultimately more persuasive reading of the post-Lisbon constitutional framework on the issue of the conclusion of non-binding international agreements is one that takes into account not only Arts 16 and 17 TEU, on the powers of the Council and Commission, but also Art. 14 TEU, in particular the Parliament’s power to exercise “political control” and “legislative functions”.
What form of parliamentary involvement does Art. 14 TEU call for? The Lisbon treaty framers envisaged a parallelism between the Parliament’s domestic legislative powers and its external powers in the area of treaty-making. Art. 218, para. 6, TFEU gives expression to this objective, in particular where it requires parliamentary consent for the conclusion of international agreements covering fields to which either the ordinary legislative procedure applies, or the special legislative procedure where consent by the European Parliament is required. Requirements of coherence as well as a more structural concern about ensuring that in the sphere of non-binding agreements as well, the two sources of democratic legitimacy of EU law-making have their part to play, call for a parliamentary consent-requirement on the basis of Art. 14 TEU that runs parallel with Art. 218, para. 6, TFEU. This implies that parliamentary consent is to be obtained whenever the Council wishes to conclude a non-binding agreement that involves a degree of policy-making in a field in which parliamentary consent is required for the adoption of domestic legislation.
In Swiss MoU, The ECJ clarifies some aspects of the constitutional framework on the conclusion of non-binding agreements by the EU (i.e. the division of labour between the Commission and the Council), but leaves others untouched (i.e. the role of the Parliament). It is to be regretted that the ECJ did not sketch out a more comprehensive picture. Perhaps the Court will have the opportunity to do so in a future case – preferably one in which not only the Commission and the Council participate, but the Parliament as well.
* PhD Candidate, Institute for European Law, University of Leuven, and Michigan Grotius Research Scholar, University of Michigan Law School, email@example.com.
 Court of Justice, judgment of 28 July 2016, case C-660/13, Council v. Commission (“Swiss MoU”).
 The Treaty of Lisbon notoriously introduced a provision (Art. 218 TFEU) setting out a detailed procedure for the negotiation and conclusion of international agreements. This provision has itself been the object of litigation before the ECJ. See e.g. Court of Justice, judgment of 14 June 2016, case C-263/14, Parliament v. Council (“Somali Pirates II”), discussed in T. Verellen, Pirates of the Gulf of Aden: the Sequel, or how the CJEU further embeds the CFSP into the EU legal order, in European Law Blog, 23 August 2016, or Court of Justice, judgment of 28 April 2015, case C-28/12, Commission v. Council (“US Air Transport Agreement”), discussed in T. Verellen, On Hybrid Decisions, Mixed Agreements and the Limits of the New Legal Order: Commission v. Council (“US Air Transport Agreement”), in Common Market Law Review, 2016, p. 741 et seq.
 Commission Decision C(2013) 6355 final of 3 October 2013 on the signature of the Addendum to the Memorandum of Understanding on a Swiss financial contribution.
 Note that under the EEC Treaty, in the 1994 case of France v. Commission, the ECJ had accepted a plea by France that the Commission lacked the power to conclude an agreement with the United States anti-trust authorities which aimed to promote cooperation and coordination and lessen the possibility or impact of differences between the EU and the US in the application of their anti-trust/competition laws. Considering the significant constitutional changes brought about by the coming into force of the Lisbon Treaty in 2009, the precedential value of this judgment is limited. See Court of Justice, judgment of 9 August 1994, case C-327/91, France v. Commission.
 The tendency of the Commission to act independently on the international stage is broader than the case discussed here. See P. Garcia Andrade, The Distribution of Powers between EU Institutions for Conducting External Affairs through Non-Binding Instruments, in European Papers, 2016, p. 122.
 Opinion of AG Sharpston delivered on 26 November 2015, case C-660/13, Council v. Commission (“Swiss MoU”). For a more elaborate discussion, see P. Garcia Andrade, The Distribution of Powers, cit., p. 115 et seq.
 Ibid., paras 57-72.
 Court of Justice, judgment of 7 October 2014, case C-399/12, Germany v. Council (“IOV”), in particular paras 56-64.
 Opinion of AG Sharpston, “Swiss MoU”, cit., paras 82-88.
 Ibid., paras 89-117.
 Ibid., paras 111-113.
 Ibid., paras 118-142.
 Ibid., para. 136.
 Ibid., paras 143-147.
 “Swiss MoU”, cit., para. 39.
 Ibid., para. 40.
 Ibid., para. 42.
 See e.g. Restatement (Third) of the Foreign Relations Law of the United States (1986), Section 301, Reporters’ note 2: “A nonbinding agreement is sometimes used in order to avoid processes required by a national constitutional system for making legally-binding agreements”.
 In the United Kingdom, for example, MoUs are routinely concluded. On the practice, see HM Government (Treaty Section, Legal Directorate, Foreign and Commonwealth Office) (UK), Treaties and Memoranda of Understanding (MOUs). Guidance on Practice and Procedures, Second Edition April 2000, updated March 2014.
 Restatement (Third) of the Foreign Relations Law of the United States (1986), Section 301, Reporters’ note 2: “[T]he political inducements to comply with such agreements may be strong and the consequences of noncompliance may sometimes be serious”.
 On this topic, see generally J. Klabbers, The Concept of Treaty in International Law, The Hague/London/Boston: Martinus Nijhoff Publishers, 1996.
 The use of effet utile-style purposive reasoning and the rejection of formalist categories by the ECJ is well recorded. For a recent discussion, see generally G. Beck, The Legal Reasoning of the Court of Justice of the EU, Oxford: Hart Publishing, 2013.
 The legitimacy of courts depends, to an important extent, on the quality and transparency of their legal reasoning, reason for which Art. 36 of the Statute of the ECJ requires that “judgments shall state the reasons on which they are based”. For a critical appraisal of the ECJ’s style of reasoning and a call for more transparency, inter alia by allowing the use of concurring and dissenting opinions, see V. Perju, Reason and Authority in the European Court of Justice, in Virginia Journal of International Law, 2009, p. 307 et seq.
 Court of Justice, judgment of 16 July 2015, case C-425/13, Commission v. Council (“Australian Greenhouse Gas Emissions”).
 Ibid., paras 85-93.
 For an example of a hypothesis in which the ECJ considers a matter to be entirely devoid of a policy-making dimension, see Court of Justice, judgment of 6 October 2015, case C-73/14, Council v. Commission (“ITLOS”), discussed by the present author in SEW, Tijdschrift voor Europees en economisch recht, 2016.
 Art. 218, para. 4, TFEU expressly provides for the establishment of a “special committee in consultation with which the negotiations must be conducted.”
 For an analysis of EU law-making processes through a “separation fo powers” lens, see e.g. K. Lenaerts, Some Reflections on the Separation of Powers in the European Community, in Common Market Law Review, 1991, p. 11 et seq.
 This point flows from the text of Treaties, which in Art. 14, para. 1, TEU empowers the Parliament to exercise “political control” and from Art. 10, para. 1, TEU, which provides that “[t]he functioning of the Union shall be founded on representative democracy”. Furthermore, in structural terms, the fact that EU law affects the legal status of individuals directly calls for a democratisation of the process through which EU law is created. The centrality of the individual in the European integration project can be traced back to Van Gend & Loos, where the ECJ held that “the Community constitutes a new legal order of international law for the benefit of which the states have limited their sovereign rights, albeit within limited fields, and the subjects of which comprise not only Member States but also their nationals” (Court of Justice, judgment of 5 February 1963, case 26/62, Van Gend & Loos, p. 12. Emphasis added). In support of a bicameral understanding of the EU legislature, see also I. Pernice, Multilevel Constitutionalism and the Crisis of Democracy in Europe, in European Constitutional Law Review, 2015, pp. 545 et seq.
 In that regard, it should be emphasised that in the present-day state of international law, the distinction between international agreements and domestic legislation is blurring, as treaties increasingly often include generally applicable rules and principles which potentially affect the legal status of individuals. In this sense, see already W. Friedman, The Changing Structure of International Law, New York: Columbia University Press, 1964, describing this development as one of a supplementing of a well-established international law of “co-existence” with an international law of “co-operation”, in which not only states, but also private corporations and private individuals are recognised as holders of rights and obligations.
 A contrario, in the absence of a policy-making dimension, neither Council nor Parliament are to be involved in the adoption of the agreement. When this will be the case arguably is an assessment that will have to be made on a case-by-case basis.