Ultra vires and constitutional identity control – apples and oranges or two drops of water?

A few days ago, Joseph Weiler and Daniel Sarmiento presented the idea of creating a New Mixed Chamber of the European Court of Justice that would deal with competence and jurisdictional conflicts in the EU. I share Weiler and Sarmiento’s concerns about jurisdictional overreach of the EU and that it needs some mitigation. I also agree that if all the highest courts of Member States were to emulate the German example, that would indeed spell the end of the EU as an integrated legal space. Finally, I agree that conflicts like Weiss, where at issue is the delineation of the jurisdictional line between the Member States and the EU, could be better addressed in an institutional setting, that would provide for the Mixed Appeal Chamber of the CJEU. In this blogpost, I would like to make an additional suggestion to Weiler and Sarmiento’s proposal and argue that, apart from ultra vires control, the New Chamber could also engage in constitutional identity review of EU law. In order to do that I will propose, what I call, the “sequential” model of adjudication on Art. 4(2) TEU, which in my opinion can be applied in the current legal setting, but which could be potentially complemented with the establishment of the new chamber.

I remember once talking to a judge sitting on a national constitutional court and asking him a question about the difference between the ultra vires and constitutional identity reviews. “They are like apples and oranges”, the judge replied. I do not share this point of view. On the contrary, I can see how one could understand identity control as an instance of ultra vires control: since the Treaties do not give the Union the competence to interfere with the constitutional identities of its Member States (quite the reverse – in art. 4(2) TEU the Union is expressly obliged to respect them), hence when the Union does so, it acts ultra vires. Having said that, I believe that the distinction should be still, at least to some point, reflected in the legal effects or consequences following these two types of control.

For an irreconcilable conflict of an EU act with a Member State’s constitutional identity should not automatically lead to its invalidation on the ultra vires (or any other) basis. A distinction (to a certain degree) between the two kinds of control might serve the EU integration process well. It is possible for an EU act (or a legal norm contained therein) to interfere with the constitutional identity of one or two Member States and yet be perfectly fine when it comes to constitutional orders of all the other Member States. Invalidating it, especially if not the act as a whole but only one or two norms contained in it were in conflict with the particular constitutional identity of a Member State, would be too excessive a mean and could open too wide the door for overruling EU legislation on the ultra vires basis. In effect, that might unnecessarily undermine the EU integration. In such a situation, I would argue that the act or the very norm, which is contrary to the Member State’s identity, should be declared by the CJEU (possibly by the new chamber) inapplicable in that particular state for the purpose of respecting its constitutional identity. At the same time, the EU and other Member States should not be forced to refrain from further integration in that regard. In such a situation the new chamber could decide on suspending the application of the act (or the particular norm) in one Member State, but only to an extent that would be necessary for the protection of its constitutional identity. Of course, that could happen only if such a suspension would not undermine the goal of the act in question. In other words, only if such an exceptional exoneration from some obligation contained in this act granted to one (or a few) Member State would not annihilate the harmonizing effect of the act in question. Otherwise, the new chamber would not be able to suspend it in the state in question. Then it could only choose between overruling the act as ultra vires or upholding it regardless of the conflict with constitutional identity of a Member State. Doing the latter it would, however, risk disobedience from a domestic court  , which would turn us back to a more political than juridical spectrum of solutions: (i) EU law amendment (an interesting idea for a special procedure in that regard, which would also recognize a special role of constitutional courts, has been recently proposed by Oliver Garner) (ii) constitutional amendment (iii) the state in question leaving the EU. Invalidating an EU act if the suspension of its application (or of some contained in it norm or norms) in a particular state was possible would, however, seem to be unnecessarily detrimental to the integration.

The sequential model of adjudication

I have already argued [see p. 509] for a similar mechanism and, in fact, I think that it could be achievable in the current legal setting via art. 267 TFEU. I called it the “sequential” rather than hierarchical model of adjudication on Art. 4(2) TEU. In this model, a Member State’s constitutional court would first issue a referral for a preliminary judgment to the CJEU backed up by (i) a clear interpretation of what and why the constitutional court considers to form part of its Member State’s constitutional identity; and (ii) a clear and open explanation as to why such identity is irreconcilable with the EU legal act or norm in question. The CJEU would then examine the request under the proportionality (most likely limited to necessity) test and decide on exonerating the given Member State from certain EU law obligations. Furthermore, the better the national request would be justified, the more limited the CJEU’s discretion should be. Thus, the sequential model of adjudication would shift the centre of gravity from the “last word” for which the constitutional courts and the CJEU seem to greedily fight for, to the “first word”. The “first word” would be what indeed should set the scene and shape further adjudicative steps (that shift is one of the key consequences of a proper – self-restrained, open and argumentatively in-depth both on the sides of the domestic courts and the CJEU – application of the “sequential” model). Of course, the CJEU could still solve the problem in other ways. First, it could overrule the EU legislation (for example if it found it incompatible with EU law of higher level – e.g. the Charter). Second, it could strike down the EU law in question on the ultra vires grounds. Third, it could also interpret it in a way, that would eliminate the conflict with the Member State’s constitutional identity. It could on the other hand reject the request, with all the risk it takes.

However, constitutional identity of a Member State should not be understood simply as its constitution and all its provisions. Rather, it should only refer to the most essential constitutional provisions. They should be decoded by the constitutional court (or the highest court) from the text of the constitution in which the pouvoir constituant expressed in the language of legal norms the national identity of this state. Since Costa and onwards it is obvious that it cannot be considered enough to simply invoke any constitutional provision to undermine EU law. Art. 4(2) TEU opens however such a possibility to some – very particular – constitutional provisions. The proposed “sequential” approach demands a lot of good faith and self-restraint on the both sides – from the domestic courts while they are decoding their constitutional identities as well as from the CJEU while it is examining the necessity to respect these identities.

So far none of the domestic courts has tried this approach. The referrals from the German or Italian constitutional courts in Gauweiler, Weiss and Taricco missed this opportunity. The issue of constitutional identity, if at all, appeared only in the reasoning – never in the questions – and was never straightforward. The cases were more of examples of a tug of war between the courts, where the domestic courts tried to take the CJEU hostage by subjecting the legality of a Union act to the review of a national court and thus gain the delusory “last word” advantage. Therefore, the CJEU has never had a chance to consider a request according to the above-mentioned “sequential” model. Yet, it already allowed certain exceptional treatment of Member States when it comes to consequences stemming from the Treaty freedoms. It did that for the purpose of respecting particular constitutional identities of those Member States (e.g. in Sayn-Wittgenstein or Bogendorff von Wolffersdorff). I realize that secondary law is much more harmonized and exceptions from it would be much more troublesome from the perspective of primacy, unity and effectiveness of EU law. Therefore, the threshold for exceptions would have to be very high. The function of Art. 4(2) TEU, however, should not be limited only to primary law. The obligation to respect constitutional identities of Member States binds Union also regarding its secondary legislation. Art. 4(2) TEU has a systemic position and general function in the EU legal order (its “de facto constitution” as Weiler and Sarmiento put it).

Adding an appeal chamber

An appeal chamber as proposed by Weiler and Sarmiento could serve well to defuse potential conflicts of the kind we are now facing after the BVerfG’s order from 5^th May 2020. That is what the authors seem to argue for as a main rationale behind introducing the new chamber, and to that I agree. It could also serve well to solve potential conflicts in graver (in the material sense) discrepancies, like the kind which is still smouldering after the Taricco saga, which did not lead to a concord, but ended on a rather worrying note after the Italian Constitutional Court showed its dissatisfaction from the perspective of Italian constitutional identity also with the Taricco II rule in its order 115/2018 (see especially paras. 5 and 11). Particularly the latter case showed a missed opportunity to apply the abovementioned “sequential” model. Had it been applied, there are different scenarios as to what could have been the outcome (unfortunately examining them would go far beyond the scope of this post). However, also in the “sequential model” the situation can still end with a discord between the CJEU and the constitutional court. A failure to reach a solution satisfying both from the perspective of EU and domestic orders, is not impossible.

Therefore, a mixed chamber could serve well as an appeal body in the “sequential” adjudication on exceptions to EU law’s uniform application for the sake of the protection of particular elements of constitutional identities of the EU Member States. It should provide another forum for solving potential conflicts between EU legislation and EU Member States’ particular constitutional identities by co-controlling the divergences in the EU. In the current legal setting, there are still some “extra-time” options available like the infringement procedure (in which the CJEU could revisit its judgment after it was opposed by domestic constitutional court). An appeal chamber as proposed by Weiler and Sarmiento, would, however, seem to be much more appropriate and legitimate (for the reasons well explained by the Authors in their post) forum for such retrial. A forum that would be ingrained in a holistically understood EU legal system and not representing isolationist domestic or federalist European perspectives.