National courts and the review of validity of EU acts after Eurobolt

If any of you are interested in EU law litigation at national courts, the Court of Justice’s judgment in Eurobolt (C-644/17) is well worth your time.

The question at stake concerns whether a national court can request information from an EU Institution in order to rule on the validity of an EU act. Needless to say, this situation arises when the EU Institution is not a party in the national proceedings, in which case the EU Institution is requested to cooperate sincerely on the sole grounds of Article 4 TEU.

In principle, this should not really be an issue, because national courts can only rule on the legality, but not the illegality of an EU act. In order for a national court to rule and confirm the legality of, say, a Commission Decision, this will be the result of a rather straight forward and prima facie analysis. In case the national court has any doubts, it must refer the case to the Court of Justice (classic Foto-Frost), and it will be for the Court of Justice to request information from the EU Institution, if necessary. Thus, these requests should typically take place in Luxembourg, not at national courts.

But in the case of Eurobolt the situation was quite different, because the applicants managed to raise a genuine doubt in the national court, but a doubt that could only be dissipated by looking at the file and the decision-making process of the challenged Council Implementing Regulation. Eurobolt submitted observations in the decision-making process of the act, but they were not brought forward to the Member States within the relevant time-period, as required by the Basic Regulation. Since Eurobolt lacked standing to bring an action of annulment against the Implementing Regulation, it had to bring an action against the national implementing acts in The Netherlands. Therefore, it made perfect sense for the Dutch Supreme Court to take a look at the Council’s file, in order to confirm that Eurobolt’s observations had not been referred to the Member States when enacting the challenged act, thus breaching the Basic Regulation.

The Court of Justice clearly stated that, indeed, national courts can request to the EU Institutions having participated in the enactment of a challenged EU act “evidence and documents” for the purpose of deciding on the validity of the contested act. The only limit that the Court introduces is the one enshrined in the Zwartveld case-law, which refers to judicial cooperation in the context of the Protocol on Privileges and Immunities. The analogy might not be fully convincing, but it’s as close as it gets to a case of cooperation between national courts and EU Institutions in gathering access to documents. As it was ruled in that judgment, the EU Institution may refuse to provide the evidence if “justified by legitimate reasons based, inter alia, on protecting the rights of third parties or the risk of an impediment to the functioning or the independence of the Union”.

Eurobolt seems quite logical and unsurprising at first, but in fact it’s a relevant decision for national litigators and courts. It is very difficult to convince a national court to review an EU act. It is also very difficult to convince them into referring the case of validity to the Court of Justice. National courts sometimes feel too overwhelmed by the idea of going into such trouble, particularly when the defendant party (usually a public authority) provides helpful arguments to convince the national court that the EU act is perfectly all right. Also, the fact that the EU Institutions are not a party in the proceedings can also be an obstacle, since most national courts don’t like being dragged into a discussion about an act whose author is absent in the proceedings.

But thanks to Eurobolt, now the national court can request the EU Institution to indirectly participate in the proceedings, by way of providing “evidence and information”. It is a subtle but effective way of inviting national courts to request the help of EU Institutions when having to rule on the validity of EU acts. In fact, once the information has been gathered, the case might take a completely unexpected turn. The parties will be given a right to submit observations in light of the evidence and information provided, and the national court will have a more realistic idea of how the decisoin-making process took place in Brussels. It’s a completely different context in which the national court can now make a determination.

In fact, Eurobolt is another turn in the process of involving EU Institutions in national proceedings, which is a positive feature that should not be an exception to the rule. In a more integrated judiciary, EU Institutions should be participating in national proceedings as a normal feature of a European court system, particularly when national courts are to decide on issues of EU law, as is the case of the review of an EU act. To this end, Article 4 TEU and the principle of sincere cooperation are proving to be useful tools that glue together the two levels of the European judiciary. In Eurobolt, the interaction happened to be very productive for the applicant: the Court of Justice sided with Eurobolt and confirmed that the Council’s Implementing Regulation suffered a procedural flaw and it was therefore null and void. If the first instance Dutch court would have had access to the file from the first place, Eurobolt wouldn’t have waited eight years for that outcome. Thus, Eurobolt is not only a nice lesson on sincere cooperation, but also on how to avoid delaying justice unnecessarily.

Should Constitutional Courts be guardians of the duty to make a preliminary reference?

Here is an interesting case for the many fans of the preliminary reference procedure.

A Constitutional Court of a Member State introduces a case-law that imposes on all courts of last instance a duty to make a preliminary reference pursuant to Article 267 TFEU as a question of constitutional principle. Thus, if a Supreme Court decides not to make a reference in breach of Cilfit or Da Costa, its judgment is now subject to a constitutional appeal for breach of the fundamental right to an effective remedy. The Constitutional Court thus becomes a guarantor of the correct enforcement of EU law by national courts.

At first sight, this is a sign of cooperation on the part of the Constitutional Court with its European counterpart, the Court of Justice. The courts of last instance in the country now have a dual incentive to make a preliminary reference when they have an obligation to do so under Article 267 TFEU. As of now, they also have a duty on constitutional grounds.

However, and here comes the catch, the Constitutional Court ruled in this way in a case in which the Supreme Court had set aside a statute due to its incompatibility with EU law. The Supreme Court did not make a reference to Luxembourg because it was convinced that the statute was in breach of a Directive. Thus, it set the statute aside on the grounds of the traditional Simmenthal doctrine.

The Constitutional Court is unhappy with this outcome, not only because the Supreme Court has refused to make a reference, but mostly because in that country it is only for the Constitutional Court to rule on the legality of parliamentary acts. The ordinary courts (including the Supreme Court) have no power to rule on that, unless they decide in a case in which EU law applies. And the Constitutional Court is unhappy with that power of ordinary courts, because it considers that its monopoly to review the legality of Parliament’s acts is being undermined.

Therefore, in its judgment upholding the constitutional appeal, the Constitutional Court rules that when a court of last instance decides to set aside a statute due to a breach of EU law, there is a reinforced duty to make a reference. And then, the Constitutional Court adds that if the court of last instance rules differently and decides not to set aside a statute, its duty to make a reference is not so qualified, and therefore the Constitutional Court will apply a low standard of review. In other words: if the Supreme Court wants to set aside a statute, almost in all cases it will have to make a reference, even if the case is acte clair or acte éclairé. But if the Supreme Court implements EU law, has doubts, but comes to the conclusion that the statute is compatible with EU law, the Constitutional Court will only review the refusal to make a reference on grounds of extreme arbitrariness or unreasonableness.

Is this approach in line with EU Law? I have my doubts.

First, is it really appropriate for a Constitutional Court to add a double layer of protection to the duty to make a reference that falls upon courts of last resort? Besides Article 267 TFEU, should a Constitutional Court introduce an additional obligation to that duty, based on the Constitution? I think that prima facie it makes sense, because it puts the Constitution at the service of EU law, facilitating the access of citizens to a rightful judge that happens to sit in Luxembourg. If the Constitution has procedural tools that guarantee access to justice, it’s appropriate that such tools apply for national and European remedies alike.

However, is it wise to discriminate between remedies on the basis of the outcome that the national court has in mind? Is it appropriate to impose on a national court a reinforced duty to make a preliminary reference only if it comes to the conclusion that a statute breaches EU law, while no such duty falls upon him (at least no reinforced constitutional duty) if it decides that the statute is perfectly ok? Isn’t there a subtle (or maybe not so subtle) invitation to courts of last instance to avoid setting aside statutes that may breach EU law? A preliminary reference takes time, it might actually not be necessary (but the constitutional duty imposes the reference anyway) and the court might simply want to terminate the proceedings. If the court walks in one direction (setting aside the statute) it will need to pass a Luxembourg test. If it walks in the other direction (the statute is fine) the case can be closed straight away, with no additional work and inquisitive testing from the Court of Justice.

The message sounds very clear to me.

It’s difficult to predict how a court would act depending on an outcome that we ignore. But be that as it may, the truth is that a Constitutional Court is imposing a dual standard for a reason that should be absolutely irrelevant for EU law. National courts of last resort should make a reference (or not) irrespective of whether the outcome is one or the other. It’s the doubt of the court on a point of interpretation of EU law that justifies the reference, not the outcome of setting aside or not setting aside a statute.

In these circumstances, I’m afraid that such a system would have a subtle but perverse effect of nudging national courts of last instance into ruling in favour of statutes, even when there are doubts as to their conformity with EU law. A sort of presumption of conformity of national law with EU law would prevail, putting a heavy and sometimes unsurmountable burden on the applicants, no matter how right they might be. And it would eventually side-line the Court of Justice in its key task of assisting national courts in finding the correct interpretation of EU law, even if it entails setting aside an act of Parliament.

The case is not the product of my imagination. It’s the outcome of the Spanish Constitutional Court’s recent judgment in case STC 37/2019, whereby it upheld a constitutional complaint on the grounds of a breach of the fundamental right to an effective legal remedy, arguing that the Supreme Court should have made a reference in a case in which it implemented EU law and came to the result that a Parliamentary statute breached a Directive. The Constitutional Court said it very clearly: the Supreme Court is misbehaving not because the acte clair or acte éclairé has been breached, but because they have set aside an act of Parliament. Thus a stricter standard of review applies, and the judgment was indeed quashed.

It’s a fascinating test case on the limits of constitutional cooperation in support of Article 267 TFEU. But it’s also a good example of how the helping hand of Constitutional Courts can be welcome at times, but only as long as they truly cooperate to ensure the proper functioning of EU law. Otherwise, it’s probably better that they keep busy minding their own business.

Crossing the Baltic Rubicon

In a classic work of American constitutionalism, Bruce Ackerman refers to the “constitutional moments” that define the political arrangements enshrined in a Constitution. In Ackerman’s work this development takes place in a paradoxical way, because such “constitutional moments” of US history do not involve an amendment of the written Constitution. They simply happen and then the Constitution changes forever, but not in writing.

EU Law has had its own share of constitutional moments. In a process characterized by integration through law, these moments have traditionally taken place in the terrain of the law and lawyers, mostly through judgments and then implemented through practice. Van Gend en Loos was a constitutional moment. The rebellion of constitutional courts against unbridled integration, starting with the German Constitutional Court in the Maastricht decision, was another constitutional moment. The discoveryof EU fundamental rights as general principles of EU law was another “constitutional moment”.

Last week another constitutional moment took place. In a rather technical area of law, the Statute of the European System of Central Banks, the Court of Justice ruled for the first time in a case that ensued in the annulment of a decision of a Member State. The Court did not declare that a Member State had failed to fulfill its obligations under EU Law. What the Court did was much more ambitious, for it annulled an act rendered by an authority of a Member State and extricated it from the domestic legal order. For the very first time, EU Law entered fully into the legal order of a Member State, declared a breach within the domestic legal order and eradicated the national legal act ipso iure. There was no need for the Member State to take any appropriate measures. No need for national courts to ensure the fulfilment of the duties enshrined in EU Law. The EU legal order did the job for them.

In the case of Rimsevics and ECB/Latvia (C-202 and 238/18), the Court was called to rule on the grounds of Article 14.2 of the Statute of the European System of Central Banks. This provision is, in the words of a dear colleague, an “extravagant rule”, unparalleled in the Treaties. It contains a remedy that looks like an infringement action, but it walks like an action of annulment. It empowers the governor of a national central bank, or the Governing Council of the ECB, to challenge before the Court of Justice the decision of a national authority to remove the governor from office if the removal has taken place disregarding the conditions established in the Treaties. When Mr. Rimsevics, Governor of the Central Bank of Latvia, was provisionally suspended from office as a result of criminal investigations, Article 14.2 of the Statute was the obvious provision to apply.

Before I continue, just a quick disclaimer: although I represented one of the parties in this case, the following lines are a purely personal account and view of the judgment. But as the reader will quickly note, it is no provocation to argue that the Court has crossed a Rubicon, a constitutional Rubicon that allows no turning back.

Contrary to the opinion of Advocate General Kokott, who purported that the Court should render a declarative judgment (in line with the requests of the applicants), the final decision was much bolder. After considering the features of Article 14.2 of the Statute, the Court, sitting in Grand Chamber, ruled that the action provided for in Article 14.2 of the Statute “has as its purpose the annulment of the decision taken to relieve a governor of a national central bank from office”.

The Court reviewed the terms and the evidence provided by the Latvian authorities and ruled that Article 14.2 had been breached. Mr. Rimsevics was removed from office without complying with the substantive requirements provided in the Statute. As a result, and here comes the constitutional moment, the Court “hereby annuls the decision of the Korupcijas novēršanas un apkarošanas birojs (Anti-Corruption Office, Latvia) of 19 February 2018 in so far as it prohibits Mr Ilmārs Rimšēvičs from performing his duties as Governor of the Central Bank of Latvia”.

In all its technicality, this rather simple statement hides a revolution with the potential of changing EU law forever. Despite its differences with public international law, EU law has remained faithful to basic principles of international orthodoxy, including the separation of legal orders that precludes international courts from judicially reviewing the legality of State action. The infringement procedure is a good example of how convention has prevailed in Luxembourg, whereby judgments are set to declare that a Member State has failed to comply with its obligations, but no annulment occurs directly in the judgment. The same happens in preliminary references in which national law is indirectly reviewed, where the Court limits its decision to state that EU law must be interpreted in the sense that it is opposed, or not opposed, to a rule of national law. The national court does the rest, including the purge or setting aside of the national act.

In Rimsevics and ECB/Latvia the Court has taken an additional step by declaring the breach and, immediately after, annulling the national act that incurred in such a breach. The annulment or repeal of national law is not left to the national authorities, but quite the contrary: it is left to the Court to purge the rule itself, irrespective of the author or position of the authority rendering the rule. If the rule would have been a constitutional provision, the Court would annul. If the rule would have been a judgment of a national court, the Court would annul. The fact that the annulled act was an individual decision of an administrative body does not deprive the judgment of its seismic impact.

One could argue that the case is limited to a very specific provision that allowed the Court to do precisely that: annul a national act. But there is nothing further from the truth: Article 14.2 of the Statute makes no reference to the annulment of a national act, it simply conveys certain features of an action of annulment, but at the same time it also shows significant traits of an infringement action. The decision to annul a national act was not evidently and unconditionally provided in a Treaty provision. It is the interpretation of the Court that has managed to exert such an extraordinary effect, unparalleled in international law and EU law itself.

The rationale underlying the Court’s decision is quite straight forward. According to the Court, “the ESCB represents a novel legal construct in EU law which brings together national institutions, namely the national central banks, and an EU institution, namely the ECB, and causes them to cooperate closely with each other, and within which a different structure and a less marked distinction between the EU legal order and national legal orders prevails” (italics added).

Thus, one could argue that this result is confined to the European System of Central Banks and no more. But think twice. “Novel legal constructs” are constantly in the making in the EU. Just take a glimpse at the recently created European Prosecutor’s Office, in which another “novel legal construct” has been put into place to safeguard the financial interests of the EU. If in the future the Court strikes down a national criminal judgment on the grounds that the European Prosecutor’s Office is a “novel legal construct” within which “a different structure and a less marked distinction between the EU legal order and national legal orders prevail”, the assertion’s premiss would be correct. The same applies to the Single Supervisory Mechanism, whereby the ECB implements national law, and national authorities enforce EU law and ECB instructions. Novel legal constructs are not the exclusive domain of monetary policy, and the challenges of European integration are blending EU and national legal orders in unprecedented ways that blur the distinction between both systems.

It would be wrong to argue that Rimsevics and ECB/Latvia is an isolated decision. Quite the contrary: by stepping into national law and purging its legal order in a way that is unparalleled in international law, the Court of Justice has finally and fully stepped into the shoes of a national court. In fact, if this judgment is seen together with the Court’s efforts to protect the independence and integrity of national judiciaries, the overall effect is one in which a newCourt has emerged. A Court sitting at a constitutional apex, assisting national courts when their integrity is undermined, confronting Member States that drift away from the rule of law, and annulling national acts when necessary, particularly in areas in which a less marked distinction between the EU legal order and national legal orders prevail.

The final portrait emerging from this description is a genuine constitutional moment, crossing a Rubicon through the northern route, via Latvia.







A tribute to Gil Carlos Rodríguez Iglesias


This morning, Gil Carlos Rodríguez Iglesias passed away. The first Spanish judge at the Court of Justice and its President from 1994 to 2003, we sadly lose today a powerful intellect that inspired many European lawyers.

Gil Carlos was the first of a generation of young Spanish lawyers that excelled in the newly democratic Spain as a professor of public international law. First, at the University Complutense of Madrid, later in Extremadura and finally in Granada, where he held the chair of public international law until his departure in 1986 to Luxembourg. I remember during my years of study in Granada, when Gil Carlos was at the time President of the Court, that the mere mention of his name was like evoking a saint or a mythical god. Despite his time far away at the Plateau du Kirchberg, he never lost his links with Granada, his disciples and his friends. The school of international law of Granada remains one of the leading centers of European law in Spain, in part thanks to Gil Carlos’ commitment and efforts.

Gil Carlos was the last President of the pre-great accession Court. A modest but mighty judicial institution of fifteen judges only, soon to be profoundly transformed by the arrival of thirteen more. His style was modest and gentle, but he leaded the Court’s works with an iron fist still remembered to this day by many who shared their time at the Court with Gil Carlos.

After his tenure at the Court he returned to Spain, where he held the chair at the University Complutense of Madrid, the school of his doctoral and post-doctoral years. He was appointed director of the Real Instituto Elcano, Spain’s leading thinktank in international affairs, and presided the Spanish section of FIDE until 2015.

Gil Carlos was a giant of European law, a force of good and reason so badly needed in these turbulent times. We will miss the man, the judge, the professor and the dignity that he exemplified, but his legacy and memory will remain, hopefully as a guiding light that helps us find the way in the unchartered waters of European integration.



The Year of the Infringement

Happy 2019! I wish all the readers of this blog a very successful and eventful new year.

And to start 2019, nothing better than a post on naughty States and the stellar role recently played by infringement actions before the Court. Who would have thought that the infringement procedure would be close to outmaneuver the preliminary reference procedure and overtake its position as the “jewel of the crown” among the remedies of the Court? But that is what we have started to witness in the past months, and will probably consolidate throughout 2019. Thus, 2019 could very well be… the year of the infringement.

This is not a completely new story. 2017 and 2018 provided extraordinary examples of how the infringement procedure could be used, particularly in the Polish forest case and, more recently, in the case of the Polish forcefully retired judges. Those cases proved that the interaction between the interim measures procedure and the infringement procedure can be crucial in order to ensure the effectiveness of EU law. The Grand Chamber ruled twice on these matters, the first time to confirm that the sanctions mechanism of Article 260 TFEU can be used in the context of interim measures (quite a ruling!), and the second to remind that interim measure inaudita parteagainst Member States can be retroactive (quite another ruling!).

But 2019 will provide us with further developments in the field of infringements, and some of them are of a huge significance, both in procedural and substantive terms.

First, the Polish proceedings concerning the independence of the judiciary will prove how far the Court of Justice is willing to go in the interpretation of a very broad and generic provision, as is Article 19 TEU, which generically ensures the independence of the judiciary. Is this provision enough to justify the launch of an infringement procedure against the reform of a national judiciary? If the Court of Justice sides with the Commission, what would happen next? Should Member States consult with the Commission any further amendments to their judiciary? Wouldn’t the EU be stepping into a closely guarded domain of national sovereignty, its domestic judiciary? It looks like the Court will probably side with the Commission, but the seismic consequences of that conclusion could be felt for many years.

Shortly after, the Court will be ruling on Commission/Hungary (C-235/17), a case that has nothing to do with the rule of law crisis in that country, but with a point that could be applied to any infringement action against any Member State: can the Commission bring an infringement action on an independent ground of review based solely on the Charter of Fundamental Rights? The case raises key issues of principle, not only about the scope of application of the Charter, but also on the Commission’s policy on infringements, which could leave the institution in a position closer to the one of an Ombudsman in the promotion and defense of fundamental rights. So far, in a thoroughly reasoned Opinion, AG Øe claims that the Commission cannot bring an action on such grounds, but the counterarguments are powerful too. Some fundamental rights are self-executing, even in horizontal situation, as the Court has stated recently in some high-profiled cases (Egenberger, Bauer…). Also, the Court’s decisions in preliminary judgments confirm that the Charter can act as a sole ground of infringement in areas only indirectly covered by EU law (think about consumer protection and the vast number of breaches of the right to an effective protection in civil judicial proceedings ruled by the Court). It would be strange if the Commission is subject to a stricter standard of proof of the link between the Charter and the implementation of EU law, simply because the case is brought through an infringement action. We shall see, but the case will condition the role of the Charter as a significant or insignificant legal instrument in the Commission’s infringement policy.

We will also hear about the unknown (to date) Article 14.2 of the Statute of the European System of Central Banks, which grants the Governing Council of the ECB the right to bring an action against a Member State having unlawfully removed a governor of a national central bank. The typical provision that was drafted with the hope that it would never be applied, finally ended up being put to the test in the case of ECB/Latvia (C-238/18), together with Rimšēvičs/Latvia (C-202/18). But is Article 14.2 an infringement procedure or an action of annulment? Or none of the above? AG Kokott has rendered a powerful Opinion arguing that the remedy looks like an infringement, moves like an infringement and talks like an infringement. But the Court will have the final say, again in Grand Chamber, in a case that will open the door to a new generation of sui generisinfringements brought by Institutions other than the Commission.

And yesterday we found AG Szpunar’s excellent Opinion in the saga of Germany’s reluctance to follow the Council’s voting instructions in the Intergovernmental Organisation for International Carriage by Rail. In yesterday’s Commission/Germany (C-620/16), an interesting procedural argument was raised by the defendant: Germany departed from the Council’s voting instructions and breached the Council Decision with which Germany disagreed. Germany took the Council to Court and challenged the Decision. Germany lost. However, in the meantime, the Commission launched infringement proceedings against Germany almost immediately after the recalcitrant vote. But the letter of formal notice and the reasoned opinion were issued once the rebellious vote had been cast, so there was nothing much to do by then. Thus, Germany argued that the Commission lacked any interest to bring the action, to which the AG has counterargued by claiming that the Commission was right to bring the action, or otherwise Member States would profit from their breaches of EU law by issuing acts of immediate effect. The AG argues, with the support of the case-law, that some breaches produce effects only in short segments of time, but that does not preclude the Commission from bringing an infringement action after the effects have faded, as long as the Commission brings the action immediately after the act takes place. This is the key: the Commission’s ability to react quickly, which it did in this case. But if the Commission had decided to await the Court’s decision in the action of annulment against the Council Decision, which could have a been a reasonable stance too, the AG’s proposal would ensue in the inadmissibility of the infringement action. We shall see what the Court finally says, but the issue is conceptually and procedurally complex, and crucial for the Commission’s infringement policy.

The infringement procedure became a rather unexciting area of practice for quite some time, mostly because the Commission did not have reasons to use it in novel ways. But the current times have put the Commission in a new role, struggling as it is now with phenomenal challenges unimagined year ago. The infringement procedure is proving to be a powerful ally in the Commission’s efforts to overcome such challenges, and so far the Court has been willing to play ball. But for how long? In 2019 we will find out.

In the meantime, happy new year to y’all.