Earlier this month, the EU-Japan Economic Partnership Agreement entered into force. It seems only yesterday that we watched Mrs Cecilia Malmström, the EU Commissioner for Trade, and Mr Fumio Kashida, the Japanese Minister for Foreign Affairs, smile at the cameras as they painted the eyes of two daruma dolls, a talisman of perseverance and good luck in Japanese culture. One could be forgiven for having missed the photogenic moment marking the conclusion of the world’s largest bilateral free trade agreement, or even its ratification by the European Parliament, in December of last year. Indeed, the negotiation, conclusion and ratification of this free trade agreement was quick and uneventful when compared to that of the EU-Canada Comprehensive Economic and Trade Agreement (CETA).

Much of this has to do with the fact that the former is an EU-only agreement whereas the latter is a so-called mixed agreement. Mixity is the legal technique of involving the Member States in the negotiation, conclusion and ratification of an international agreement alongside the Union. It is a common feature of EU foreign affairs because there are limits to the EU’s (external) competence and the envisaged agreement may exceed those. Since competences not explicitly or implicitly conferred on the EU by the Treaties remain with the Member States (Article 5(2) TEU), the latter must frequently be involved to supplement EU competence. Thus, when asked about the competence to conclude the EU-Singapore free trade agreement in Opinion 2/15, the Court held that the EU had exclusive competence over foreign direct investment but the Member States needed to be included with respect to the parts of the agreement that related to portfolio investment and investor-state dispute settlement.

While attractive from a Member State perspective, mixity diminishes the efficacy of EU external representation. Negotiations are more cumbersome with that many actors around the table and EU negotiating power is weakened because a large trading partner could put pressure on an individual Member State to obtain concessions from the EU. Securing approval is also much more difficult: a mixed agreement requires agreement in the Council and consent of the European Parliament (pursuant to Article 218 TFEU) but also approval by all Member States and ratification in accordance with their own constitutional requirements. This explains why, in autumn 2016, the Walloon parliament, by refusing to give the Belgian government permission to sign CETA, could hold the entire EU hostage. So, while mixity strengthens democratic control and is intimately tied up with the nature of the EU as a divided-power system, it is often portrayed as a bug rather than a feature.

For this reason, some argue that recourse to mixity is constitutionally permissible only when objectively necessary. Advocates of this view – most notably Advocate General Kokott – maintain that mixity is merely ‘facultative’, that is, dispensable, when an envisaged agreement falls squarely within an area of external competence shared between the Union and its Member States. In this scenario, there is nothing, barring special circumstances, that keeps the Union from making use of that competence. To the extent that it does, Member States lose the right to exercise their competence in the area (per Article 2(2) TFEU) and therefore their right to participate in the negotiation, signing and conclusion of the agreement. Given that the Union can act alone, it should. Theoretically, the Council could decide to exercise its shared competence only partly, leaving room for the Member States’ involvement alongside the Union. Advocate General Kokott, for one, has rejected that argument, submitting that the Council could not voluntarily waive some or all of its powers, especially when this would weaken the effectiveness and coherence of EU external representation. More recently, however, she asked the Court to hold that there should be, at the very least, objective evidence amenable to judicial review of the Council’s decision to refrain from making full use of its competence.

The Court has kept off the boat on condemning facultative mixity. It took years for a judgment to confirm the preliminary point that shared competence and mixity were altogether different issues. As recently as Opinion 2/15, Member States’ participation seemed to flow automatically from the fact that competence with respect to (part of) the envisaged agreement was shared in nature. However, in its 2017 judgment in Germany v. Commission (‘COTIF’) the Court took pains to clarify that it had simply acknowledged that there was no political majority in the Council to exercise EU shared competence. Be that as it may, as a practical matter, the choice for mixity remained a political decision, not constrained by legal rules apart from decision-making procedure.

Pressed on the issue, the Court held in its recent Grand Chamber judgment in Commission v. Council (‘Antarctica’) that there may be compelling reasons to involve the Member States alongside the Union even when the envisaged external action falls squarely within shared competence. The case concerned the validity of two Council decisions approving, among other things, the submission of a reflection paper on the creation of a marine protected area to the Commission for the Conservation of Antarctic Marine Living Resources (‘CCAMLR’) on behalf of the Union and its Member States. The Commission argued that the Council should have submitted the document on behalf of the Union alone because the matter fell within an exclusive EU competence or, at the very least, a competence that was shared. The Court, however, found the creation of a marine protected area in the Antarctic to be primarily about protecting the environment — a shared competence under Article 4(2)(e) TFEU — and saw no proof of supervening ‘ERTA-style’ exclusivity under Article 3(2) TFEU. As a result, the Court was in an ideal position to clarify whether shared competence precluded mixity, i.e. whether facultative mixity was constitutional.

Yet, in a surprising turn, the Grand Chamber held that in this specific case international law compelled the Member States’ participation. The CAMLR Convention explicitly permits regional economic integration organisations like the EU to accede, to become members of the CCAMLR (Article XXIX), and to participate in the taking of decisions in lieu of its Member States (Article XII). Yet, to the Court, that treaty does not allow the Union fully to replace its Member States. Such exclusion would undermine the special rights and responsibilities of those Member States that are not just members of the CCAMLR but also have the status of Consultative Party under the older and more general Antarctic Treaty. It would undermine the ‘coherence’ of the Antarctic treaty system.

The Antarctica judgment is a product of the Court’s reluctance to confront the issue of facultative mixity head on. On a literal interpretation, Article XII(3) of the CAMLR Convention allows a regional economic integration organisation to participate in the taking of any decision without its Member States. The Convention encourages CCAMLR members to accede to the Antarctic Treaty but, conversely, does not insist that CCAMLR members who are both consultative parties to the Antarctic Treaty and Member States of a regional economic integration organisation (which itself is also a CCAMLR member) must always participate in that body’s decision-making. It looks like the Court went out of its way to provide reasons for Member States’ participation.

That being said, the judgment also establishes that just because the Union can theoretically act alone from a competence point of view, does not mean that Member States’ involvement cannot be justified. Thus, the Court’s recourse to systemic interpretation in an attempt to minimise potential threats to the integrity of the Antarctic treaty system may be viewed as an expression of the EU’s mission to ‘contribute to […] the strict observance and the development of international law’ (Article 3(5) TEU), even if this comes at a cost for the unity of external action. More likely, however, is that here Member States’ participation benefits the efficacy of EU external action. Article XII(4) of the CAMLR Convention, by stipulating that regional economic integration organisations ‘shall have only one vote’ when they take decisions in the CCAMLR, suggests that the Union’s voting power would diminish drastically if it decided to act alone. If this consideration played a role, the Court found mixity not just lawful but, as Christina Eckes put it, desirable.