Chaos averted or executive overreach?

No “No-Deal Brexit”! On Christmas Eve 2020, a continent breathes a sigh of relief. The EU and the UK adopt the “EU-UK Trade and Cooperation Agreement” (TCA) to resolve years of intense controversy.

But how could the TCA prevent the chaos of a “No-Deal Brexit” before even entering into force? After all, ratification of the TCA required consent of the UK Parliament and the European Parliament (EP). And while the UK Parliament adopted a last-minute implementation bill on 30 December 2020, the EP insisted that it would not vote on any agreement reached after 21 December 2020. The solution was provisional application of the TCA, based on a decision by the Council of the European Union (Council).

So, is the instrument of provisional application an easy “silver bullet” that saved Europe from a “No-Deal Brexit”? As a matter of international law, provisional application creates hard obligations to implement the TCA. This results in inevitable tensions with the constitutional separation of powers, which the ratification procedure is supposed to preserve. Accordingly, the potential success story of a “No-Deal Brexit” averted at the last moment should not be misunderstood as a blank check for provisional application of treaties.

What is provisional application?

Provisional application is a tool to give effect to a treaty before the parties conclude the often lengthy ratification procedure. States and international organizations use this tool not only to avert imminent crises, but also for more “indefinite” arrangements where ratification is particularly complex. A recent example of the latter category is the Comprehensive Economic and Trade Agreement (CETA) between the EU and Canada, which has been provisionally applied since 2017.

The (very) basic elements of provisional application are codified in the respective Articles 25 of the 1969 Vienna Convention on the Law of Treaties (VCLT) and its 1986 counterpart concerning international organizations (VCLTIO). As the EU is not a State and not party to the VCLTIO, neither convention applies to the TCA. Accordingly, references to provisions of these Conventions should be understood to refer to customary rules reflected therein.

According to Articles 25 VCLT and VCLTIO, States and international organizations can agree to provisionally apply a treaty, pending its entry into force. Unless the treaty otherwise provides or the parties have otherwise agreed, provisional application may be terminated by notification of the intention not to become a party to the treaty.

One essential question is not addressed in these provisions and has been subject to controversial debate in the past: Does the agreement to provisionally apply a treaty create binding obligations? By now, a growing international consensus suggests it does. In 2018, the International Law Commission concluded that provisional application

“produces a legally binding obligation to apply the treaty or a part thereof as if the treaty were in force […] unless the treaty provides otherwise or it is otherwise agreed” (here at 203-204).

States commenting on this draft generally accepted the binding nature of provisional application (see here at 21-23 and here at para. 95). High profile cases in practice have demonstrated the potentially far-reaching legal consequences of provisional application: For example, Russia is currently attempting to avert enforcement of $ 57 billion in damages awarded by an Investor-State Dispute Settlement Tribunal established under the European Energy Charter – a treaty Russia had provisionally applied, but not ratified.

Obligation to provisionally apply the TCA

One way to agree to provisional application is to include a provision to that effect in a treaty (Article 25(1)(a) VCLT). According to Article FINPROV.11(2), the TCA provisionally applies from 1 January 2021

“provided that prior to that date they have notified each other that their respective internal requirements and procedures necessary for provisional application have been completed”.

After the Council decided to agree to provisional application and the UK Parliament adopted the TCA implementation bill, that exchange happened by 31 December 2020. Since the TCA does not restrict the scope of provisional application, the EU and UK are under an obligation to apply the entire agreement from 1 January 2021.

When does provisional application end?

Concerning the end of provisional application, the TCA provides in Article FINPROV.11(2):

Provisional application shall cease on one of the following dates, whichever is the earliest:

(a) 28 February 2021 or another date as decided by the Partnership Council; or

(b) the day [of the entry into force of the Agreement].

The “Partnership Council” is a so-called “Treaty Body” established under the TCA. It comprises representatives of the parties and is endowed with various competences concerning implementation, application, interpretation, and even amendment of the TCA (Article INST.1). Decisions of the Partnership Council are binding and adopted by mutual consent (Article INST.4).

A reasonable reading of Article FINPROV.11(2) suggests that the Partnership Council has the power to extend the 28 February deadline of provisional application. It would be peculiar if the “earliest” date necessarily referred to both alternatives under subparagraph (a), meaning that the Partnership Council could only shorten the 28 February deadline. This clause may soon become relevant as the EP already hinted at proposals to “slightly” extend the period of provisional application to consider the TCA during the March plenary session.

The “constitutional” perspective: Did the Council act ultra vires by agreeing to provisional application?

As a matter of international law, the Council has put the EU under an obligation to apply the entire TCA. Even though this obligation is (for now) limited to a short period of time until 28 February 2021, one might question whether the Council had the unilateral power to do so.

At the outset, the text of Article 218(5) TFEU quite clearly states that for agreements with third parties the Council “shall adopt a decision authorising the signing of the agreement and, if necessary, its provisional application before entry into force.” Consent of the EP is necessary only before a decision “concluding” (i.e., ratifying) certain agreements under Article 218(6)(a) TFEU, which applies to the TCA.

However, there are three particular points of debate which could call the power of the Council into question:

First, some commentators argue that legitimizing the far-reaching legal effects of “mere” provisional application requires EP consent. This analogous  application of Article 218(6)(a) TFEU to the stage of provisional application faces the obstacle that Article 218(10) TFEU prescribes a comprehensive right of the EP to be fully informed “at all stages of the procedure“, which suggests that there is no unintentional gap in the framework of Article 218 TFEU. The EP, for its part, has in recent practice successfully insisted on being consulted before decisions on the provisional application of trade agreements (for an overview see here and here). Not so in the face of the impending “No-Deal Brexit”. On 28 December 2020, the EP’s Conference of Presidents released a rather political statement accepting provisional application given the “particular, unique and specific circumstances” of the negotiation process while emphasizing that this “neither constitutes a precedent nor reopens established commitments made among EU institutions”.

Second, provisional application of agreements which affect competences of both the EU and its Member States (so-called “mixed agreements”) may require consent of Member States. This was effectively the position taken by the German Constitutional Court when it insisted that provisional application of CETA based on a Council decision be limited to those parts of the Agreement which do not affect competences of Member States (here at paras. 67-70). So far, claims that the TCA affects the competences of Member States seem not to have surfaced. However, untangling EU and Member State competences in treaty-making remains complex (see, for example, here and here). If scrutiny of the 1,200+ pages of substantive trade and cooperation provisions raised concerns in that regard, this could have implications for the legality of provisional application under EU law.

Third, the practice of the EU to establish powerful “Treaty Bodies” such as the TCA Partnership Council has attracted some criticism. The German Constitutional Court labelled the democratic legitimation of a similar Treaty Body – the CETA Joint Committee – “uncertain” (here at para. 65). The democratic legitimation of the TCA Partnership Council is certainly not superior, especially considering that the EP had consented to provisional application of CETA. The actual decision-making process in the Partnership Council is not likely to add democratic legitimation either. As usual, the Council decided that the Commission will represent the EU based on positions taken by the Council in accordance with Article 218(9) TFEU (here at (7) and Article 2). Concerning the role of the EP, Article 2(3) of the Council Decision only provides a vague disclaimer that “[t]he European Parliament shall be put in a position to exercise fully its institutional prerogatives throughout the process in accordance with the Treaties.“  

Could an ultra vires act lead to the invalidation of provisional application?

Given these concerns over the Council’s competency, could the provisional application of the TCA be invalidated?

At the outset, it is a fundamental principle that a violation of internal law provides no escape from treaty obligations (reflected in Article 27 VCLT). A narrow exception from this principle is codified in Articles 46 VCLT and VCLTIO (whose applicability in the context of provisional application  is not uncontroversial). According to these provisions, States and international organizations can only invoke a violation of internal provisions of competence to invalidate their expressions of consent if the violation was “manifest” and concerned an internal rule of “fundamental importance”.  A “manifest” violation means one that is “objectively evident”.

Since arguments that the Council may have acted ultra vires are far from “objectively evident”, the EU could not potentially invalidate its consent to provisional application of the TCA.

Evaluation

The case of the TCA has the potential to become a successful example of chaos averted through provisional application of treaties. If ratification proceeds smoothly, most of the issues raised above will remain theoretical. This example might even challenge the notion that it is provisional application of treaties and not just ordinary political pressures which present parliaments with a fait accompli. After all, provisional application gives the EP an opportunity to thoroughly scrutinize the agreement before taking a vote. Was the UK Parliament really in a better position when it adopted a last-minute implementation bill?

However, the TCA may also become an example of executive overreach. If ratification falters, the prospect of another “No-Deal Cliff” would make extension(s) of provisional application through the Partnership Council a likely option. And besides the 28 February deadline, the TCA provides no limitations on provisional application to safeguard the separation of powers. Admittedly, expecting an intricate provision which strikes a balance between such safeguards and the effective operation of the TCA may be unrealistic for a last-minute agreement. Nevertheless, it is important to highlight that the more significant the obligations drafters choose to create by way of provisional application, the stronger the argument for some prior parliamentary approval. A deadline or the possibility of termination are important factors, but they do not prevent potentially far-reaching legal consequences from arising in the meantime.

In any event, conclusions drawn from the provisional application of the TCA should consider that this is an exceptional case – not just due to the impending “No-Deal Chaos”, but also because the TCA reshapes a pre-existing close relationship and does not create a new one from scratch. Against this backdrop, it is difficult to disagree with the EP’s Conference of Presidents caution that provisional application of the TCA “should not serve as a blueprint“ for the future.