Self-amending Treaty rules and the birth of constitutional robotics

Among the many fascinating questions that Brexit has brought to the legal arena, there is one that has not raised much attention: the need to repeal the references to the United Kingdom in primary law, mostly the Treaty on European Union and the Treaty on the Functioning of the European Union. Once the United Kingdom is gone, its presence in the Treaties is a piece of legal archaeology that must be removed. There are no possible compromises here.

This is a purely technical adjustment that should raise no concerns. But unfortunately it is not as simple as it may seem.

The obvious choice would be to introduce minor adjustments to the Treaties in the withdrawal agreement. The treaty establishing the conditions of Brexit should be the instrument that cleans the Treaties of all references to the United Kingdom. However, this is not possible due to the withdrawal agreement’s bilateral nature and parties: the EU is party to the withdrawal treaty, not the Member States. The EU cannot amend multilateral treaties entered into by Member States (i.e., the TEU and the TFEU) and, above all, it cannot amend its constitutive Treaties, only its Member States can.

Another option is to enact an ad hoc multilateral treaty, for example the treaty in which a transitory agreement UK/EU would be envisaged to ensure the transit from Brexit towards a future trade agreement. This option is not workable for several reasons: the United Kingdom cannot be a party to this agreement, because it cannot participate in an agreement amending Treaties to which it is not a party. But above all, an international treaty amending the TEU and the TFEU will find the insurmountable obstacle of the Defrenne II case-law, which is now enshrined in the Treaties themselves: all amendments to the Treaties must be undergone through the Treaty amendment rules to this effect. In other words, all Treaty change must pass through the procedures of Article 48 TEU. The only exceptions, and subject to very strict conditions under Article 49 TEU, are Treaty “adjustments” in accession treaties. But that’s all. And this is not an accession of a new Member State, but exactly the opposite.

It therefore appears that the only option left for a rather simple reform is a Treaty amendment by way of Article 48 TEU. It is tempting to assume that such a modest reform would be an excellent candidate for a simplified revision procedure. But alas, EU was never an easy travelling partner and the simplified revision procedure is available only for amendments to Part Three of the TFEU. The references to the United Kingdom are everywhere throughout the Treaties, so the simplified revision procedure will not do (in fact, the references are everywhere except in Part Three of the TFEU). We are thus in the hands of the ordinary revision procedure, which requires ratification by all Member States.

Ratification in all Member States might seem like the least terrible option, but it raises a question of timing. If an agreement in reached between the United Kingdom and the EU on the conditions of withdrawal, the ratification process will require the consent of the British Parliament and the EU (Council and European Parliament). This will be a rather straightforward process (fingers crossed), but definitely a much quicker one that going through twenty-seven national ratifications. If the withdrawal agreement needs to be synchronized with the ordinary revision for the amendment of the Treaties, the former will be de facto conditioned to national ratification processes in all Member States. This might take a long time. Or it might risk Brexit becoming hostage of a national (or regional) Parliament. Nobody will really care if the revision only concerns very straightforward amendments to a few references to the United Kingdom. National politics are politics, and any chance is a good chance for small parties to make noise and threat with blowing up the entire Brexit process.

And thus we arrive at the brave new world of robotics and self-amending Treaty rules.

My suggestion is that scrubbing all references to the United Kingdom needs no revision procedure at all. It is an automated effect of the triggering of Article 50 TEU. Article 50 TEU provides a procedure for withdrawal, but it also contains a Treaty-amending rule for the indispensible “adjustments” resulting from the withdrawal of a Member State. This is an automated amendment that does not allow for any further changes in the Treaties. All it can do is “clean” the Treaties and erase all traces of existence of the withdrawing State. It is automatic, the mere effect of legal robotics.

The self-amending virtues of Article 50 TEU are not only practical, they are also crucial to preserve the integrity of EU Law, particularly after in the case of a hard Brexit with no withdrawal agreement. The options I mentioned earlier rely on the existence of an agreement between the United Kingdom and the EU. But if there is no agreement, the withdrawal will take place anyway and the Treaties will need an immediate filtering of its rules. The only way to ensure an efficient withdrawal, including the removal of references to the withdrawing State from the Treaties, is by providing an automated amendment mechanism.

This feature of Article 50 TEU is also coherent with the automated guillotine approach to the timing of the withdrawal process. A Treaty rule establishing that “the Treaties shall cease to apply” to the withdrawing State after two years, is anticipating an amendment to the Treaties. The fact that Article 50 TEU makes an explicit reference to “the Treaties” reinforces the idea that the said Treaties undergo a transformation. Therefore, the guillotine rule is not only a tool for political pressure in negotiations, it is also a rule governing the legal effects of a clear-cut withdrawal. Such withdrawal entails a revision of the Treaties, an automated revision, just as automated as the running two-year clock.

Such automated revision is necessary in order to ensure the proper functioning of the Union, particularly when the withdrawing Member State has been so talented in scratching special arrangements that are now scattered throughout the Treaties. In the case of the withdrawal of a country like the United Kingdom, the amendment of the Treaties is not just a surgical act of removal of a specific reference, but a broad revision of the practical totality of Treaty chapters. The Preambles of the TEU and TFEU contain references to the United Kingdom. Part Four of the TFEU, on overseas countries and territories, makes very relevant references to the non-European countries and territories with special relations with the United Kingdom. Article 355 TFEU, on the territorial scope of the Treaties, refers several times to the United Kingdom. The Protocols specify aspects as varied as the United Kingdom’s capital subscriptions in the European Investment Bank (Article 4, Protocol 5), or its arrangements under the Schengen acquis (Protocol 19). Some Protocols are specifically concerned with the enactment of rules for the United Kingdom only (Protocol 15 on the United Kingdom’s opt-out from the euro) or for the United Kingdom bilaterally with another Member State (Protocol 20 on the application of Article 20 TFEU to the United Kingdom and Ireland). On the day after Brexit, these provisions must be gone, or otherwise a lot of uncertainty will be created.

It could be argued that the automated revision resulting from Article 50 TEU would only occur in case of a hard Brexit. If there is an agreement, there will also be a broad Treaty amendment that would reach beyond the mere cleansing of references to the United Kingdom. For example, an agreement on the UK/Irish border will probably require specific recognition in the Treaties, providing for special rules on border control that might derogate from Treaty rules. This agreement could provisionally enter into force before ratification, and any revision of references to the United Kingdom could be included there. However, this option raises concerns once again. It is not clear if Article 48 TEU and, in particular, the ordinary revision procedure allows for temporary entries into force of Treaty amendments. And above all, it is dangerous to mix specific arrangements for the purpose of bilateral relations with a broad cleansing of the TEU and the TFEU. The chances of collision with the orthodoxy on Treaty amendments and with the preservation of the autonomy of EU Law are a risk too serious to be taken.

I therefore suggest that the way forward, whether there is a withdrawal agreement or not, is to assume that Article 50 TEU provides an automated removal of all references to the United Kingdom in the Treaties, including its Protocols and Declarations. The day after Brexit, the Publications Office should provide a new consolidated version of the Treaties in which all references to the United Kingdom disappear. If this causes any disarray, it will have to be handled either through secondary law in a preventive way, or afterwards through an ordinary revision procedure with ratification in all Member States. But that is the price to pay for a clean rupture in Law, and not a messy faux Brexit, real in practice, false in Law.

After all these years, we were unaware that there was an additional Treaty amendment procedure in the Treaties. In fact, there were more procedures than those envisaged in Article 48 TEU, because Article 49 TEU allows for specific “adjustments” in accession Treaties. What is new is the ability of Article 50 TEU to act as a Treaty amendment rule as well, in a fashion that is different to other procedures, but coherent with the dynamics of the time-bomb philosophy of the withdrawal process. After all, an automated withdrawal within a fixed time-period also needs automated Treaty amendments.

Welcome everybody to the age of constitutional robotics.