‘Conditional’ Primacy of EU Law and Its Deliberative Value: An Imperfect Illustration from Taricco II

By Michał Krajewski

To err is human and so it is with judges, even the highest ones. Take the long awaited ECJ’s judgment in case C-42/17, M.A.S. & M.B. (Taricco II). This is already a second ruling on the Italian statutes of limitation applicable to pending criminal proceedings regarding VAT fraud. The statutes of limitations turned out too short for the Italian justice system, facing workload and efficiency problems. As a result, a significant number of persons guilty of serious VAT fraud might go unpunished. This in turn would undermine the effective protection of the financial interests of the EU (Article 325 TFEU). Previously, in case C-105/14, Taricco I, the ECJ had obliged Italian criminal courts to disapply the statutes of limitations in VAT cases, in order to give full effect to Article 325 TFEU. However, following the firm opposition from the Italian Constitutional Court (the ‘ICC’), the ECJ revoked the said obligation in Taricco II.

In this blog post, I will point to ambiguities in the ECJ’s reasoning in Taricco II and to further problems that this ruling may generate. I will argue, however, that the shortcomings should not overshadow the generally positive conclusion that we may draw from the Taricco saga. In my view, this saga illustrates a positive side to the ‘conditional’ acceptance of EU law primacy by national constitutional courts as the latter provide checks and balances on the ECJ’s enormous judicial power. By threatening to disapply EU provisions, they can force the ECJ to seriously engage in a deliberative process, eventually leading to the correction of mistakes that the ECJ will surely commit from time to time.

The Taricco saga so far

The Taricco saga has generated a heated debate among EU law scholars. For the discussion of the Taricco I case, see the post by Lasalle on this blog. In my previous post on this blog, I argued that the ECJ’s judgment in Taricco I is questionable under EU law itself. The ECJ omitted in its reasoning the EU general principle of legal certainty. Moreover, its decision adversely affected the fundamental right to a court hearing within a reasonable time. What the ECJ did was to oblige national courts to worsen significantly the legal position of persons accused of VAT fraud but still presumed to be innocent, if otherwise, ‘the imposition of effective and dissuasive penalties in a significant number of cases of serious fraud’ would be prevented. By imposing such a vague criterion, the ECJ made Italian courts to consider, while deciding individual cases, the systemic issues of criminal policy. Thus, national courts were supposed to replace the Italian legislature.

In another post published on this blog, Ziccardi Capaldo spoke against framing the legal problem in the Taricco saga as a clash of the protection of Union’s financial interest and fundamental rights. In her view, that serious fraud is a crime against the fundamental rights and interests of citizens. From this perspective, the ECJ aimed in Taricco I to protect the fundamental rights and interests of EU citizens. I think there are two important points that should be raised here (see also the reply by Gill-Pedro).

Firstly, whenever legal principles (such as the protection of Union’s financial interests and legal certainty) are in conflict, they must be balanced in such a way as to keep at least the minimum of each. This mode of balancing is prescribed by Article 52(1) CFR which stipulates that ‘any limitation on the exercise of the rights and freedoms recognised by this Charter must (…) respect the essence of those rights and freedoms’. The origin of the problem in Taricco I was that the ECJ omitted the issue of how national courts should protect the fundamental rights belonging to the persons charged with VAT fraud. It confined itself to state in para. 53 that these rights must be protected by national courts somehow. But at the same time, we could infer from the operative part of the Taricco I judgment that the protection of the Union’s financial interests enjoys absolute priority. Hence, even if indeed the ECJ in Taricco I aimed to protect the rights and interests of EU citizens, it surely failed to consider properly the rights of the EU citizens most directly concerned by its ruling.

Secondly, at any rate the solution reached in Taricco I interferes excessively with the fundamental rights of persons charged with VAT fraud. Indeed, the rights and interests of the victims of such fraud could be perfectly protected without infringing the general principle of legal certainty and the fundamental rights of the accused. The relevant decision to modify the statutes of limitation should be taken by the Italian legislature through the adoption of a generally applicable statute, and not by courts deciding on a case by case basis, without any clear legislative guidance. Furthermore, as I claimed in my previous post, it is arguable that EU law allows for a retroactive extension of limitation periods, but only if such an extension is limited and proportionate.

Reconstructing the ECJ’s reasoning in Taricco II

The ECJ acknowledged its mistake. Yet, instead of creating a dangerous precedent by authorising an exception to the primacy of EU law, it modified its previous stance. It elaborated further on the obligation of national courts to protect fundamental rights and left the final decision to the ICC. However, the problem with Taricco II is that the ECJ’s reasoning is extremely difficult to follow (as pointed out also by Bassini and Polinico and Burchardt). Below, I will present my reading of this judgment.

Da mihi factum, dabo tibi ius

The ECJ put the blame for its mistake on the national court that referred the Taricco I case (Taricco II judgment, paras 22-28). This court did not fulfil its obligation to set out all the relevant matters of fact and national law. The ECJ reminded that its role is to provide the referring court with the interpretation of EU law which it deems applicable within the factual and legal context as described by the referring court (para. 24). According to the ECJ, certain issues had not been ‘drawn to its attention’ previously (para. 28). What the ECJ meant was that under Italian law ‘the requirements of foreseeability, precision and non-retroactivity inherent in the principle that offences and penalties must be defined by law apply also (…) to the limitation rules for criminal offences relating to VAT’ (para. 58, see also para. 45). One may have doubts whether this is really a particularity of Italian law (see my previous remarks on the EU general principle of legal certainty). What is sure is that the ECJ found thus an excuse to provide a different assessment of the situation that arose in Italy.

Legal certainty as a limit to direct applicability

How did this ‘new’ information impact the Court’s assessment? On the one hand, it did not change anything with regard to the incompatibility of the Italian legislation with Article 325 TFEU. The ECJ confirmed that this legislation need to be amended but this time it specified that this must be done by the national legislature (paras 42 and 61). One the other hand, the ECJ changed its view on whether also national courts bear the responsibility for giving full effect to Article 325 TFEU.

To begin with, the ECJ reconstructed the ‘requirements of foreseeability, precision and non-retroactivity’ of law that are ‘inherent in the principle that offences and penalties must be defined by law’ (para. 58). These requirements form part of EU law (para. 51). They stem, at the same time, from Article 49 of the EU Charter of Fundamental Rights (Charter), the constitutional traditions of Member States and the European Convention of Human Rights (paras 52-57). Although those ‘requirements’ are inherent in Article 49 Charter, the ECJ treats them as independent EU legal norms, distinct from this provision. Arguably, Article 49 Charter is a further concretisation of the requirements in the field of criminal law.

The ECJ noted that ‘the requirements (…) apply also, in the Italian legal system, to the limitation rules for criminal offences relating to VAT’ (para. 58). It follows that when the persons concerned allegedly committed VAT offences, they could legitimately expect that the statutes of limitation periods in force at that time will remain unchanged as concerns them. But the ‘requirements of foreseeability, precision and non-retroactivity’ are also recognised in the EU legal order (para. 51), although in their EU version they do not apply to statutes of limitation. Nevertheless, if a national criminal court disapplied the statutes of limitation to enable the imposition of penalties for VAT fraud as required by the Article 325 TFEU, it could lead ‘to a situation of uncertainty in the Italian legal system as regards the determination of the applicable limitation rules, which would be in breach of the principle that the applicable law must be precise’ (para. 59). This is because Italian constitutional law raised ‘legitimate expectations’ with regard to the stability of the statutes of limitation applicable to pending criminal proceedings concerning VAT fraud. The ‘legitimate expectations’ must also be protected under the EU version of the ‘requirements’ (para. 59; it may be disputed, however, whether the ECJ talks here about the EU or national version of requirements, see the next paragraph). Hence, if national courts confirm that the legitimate expectations arose, not only are they not obliged to disapply the statutes of limitation (para. 59); they are ‘precluded’ from doing so in view of the ‘requirements’, arguably, in their EU version (para. 60, referring to para. 58).

Another way of reading Taricco II has been presented by Burchardt. She claimed that the ‘de facto point of reference is domestic constitutional law and its understanding of the principle of legality’. In her view, the ECJ relied on the ‘requirements’ in their national and not their EU version. If this is correct, it means that the ECJ applied directly or ‘approved’ the ICC’s interpretation of Italian constitutional law. I admit that the textual interpretation of the judgment – of which structure and text are so unclear and full of ambiguities – does not allow to decide whether Burchardt’s or the present interpretation is the correct one. I chose the present one because I find it next to impossible that the ECJ decide to assume the power to confirm or approve the interpretation of national constitutional law.

What is sure, though, is that the ECJ introduced a temporal limitation of the obligation to apply the statutes of limitations in spite of Article 325 TFEU. The statutes of limitation should still be applied but only to the persons accused of committing VAT infringements before the delivery of the Taricco I judgment (para. 60). This seems logical as after the delivery of the Taricco I judgment, anyone should be able to reasonably predict the effects of Article 325 TFEU (ignorantia iuris nocet).

A preliminary list of doubts

One could make a long list of doubts and reservations to the clarity of ECJ’s reasoning.

Firstly, why does the ECJ talks about the ‘requirements of foreseeability, precision and non-retroactivity’ of law, that are in its view ‘inherent’ in the principle of legality instead of referring to the well-known general principle of legal certainty and legitimate expectations? What is the added value of the new concept of ‘requirements’?

Secondly, did the ECJ rely on the Melloni doctrine (discussed previously on this blog by Franssen)? Under the Melloni doctrine, ‘national authorities and courts remain free to apply national standards of protection of fundamental rights, provided that the level of protection provided for by the Charter, as interpreted by the Court, and the primacy, unity and effectiveness of EU law are not thereby compromised’ (this doctrine has been applied in C‑168/13 PPU, Jeremy F). Admittedly, the ECJ uses this formula in para. 47 but it cites C-617/10, Åkerberg Fransson (discussed by Ankersmit) and not Melloni. Neither did it mentioned Article 53 Charter on which Melloni relies. Moreover, one can hardly see how the condition for applying higher national standards, included in Melloni formula – i.e. that the primacy, unity and effectiveness of EU law must not be compromised – are fulfilled in light of the ECJ’s findings that the Italian statutes of limitation compromise the protection of Union’s financial interest and still need to be changed by the legislature (see also on similarities and dissimilarities between Melloni and Taricco II: Burchardt, Bassini and Polinico and Sarmiento).

I believe there are more arguments against interpreting Taricco II as an application of the Melloni doctrine. Under the Melloni doctrine, if EU law does not provide the full harmonisation of a matter, national criminal courts ‘are free’ – i.e. they are permitted but not obliged under EU law – to apply higher national standards of fundamental rights. That is, the obligation to do so stems from national constitutional law, EU law merely permits it. Accordingly, in para. 59, the ECJ held the national court ‘is not obliged’ to disapply the statutes of limitations. In other words, national courts are permitted to apply the statutes of limitations, in accordance with the requirements stemming from its national constitutional law. This points to Melloni. But immediately in the next paragraph, the ECJ held that the requirement of precision ‘precludes’ national courts from disapplying the statutes of limitations. In other words, national courts are obliged to apply the statutes of limitations. This excludes the interpretation of Taricco II as an application of the Melloni doctrine because the latter provides only for a permission and not an obligation to apply higher national standards of fundamental rights, if the matter is not fully harmonised by EU law. If eventually the ECJ confirmed the national courts’ obligation to apply the statutes of limitations stemming from the ‘requirements’ (para. 60), such an obligation and the ‘requirements’ themselves must stem from EU law. Otherwise, it would mean that the ECJ assumed the power to interpret or at least confirm the interpretation of national constitutional law (as claimed by Burchardt). But even in such a case, Taricco II would represent a new doctrine, different from Melloni.

Thirdly, what is the legal character of national statutes of limitations once they are harmonised by EU law? The ECJ upheld its view that under EU law, by default, the principle of legality does not apply to statutes of limitations. At the same time, a partial harmonisation of statutes of limitations applicable to criminal proceedings relating to VAT has been carried out, as the ECJ observed (para. 44), by the Directive 2017/1371/EU of the European Parliament and the Council of 5 July 2017 on the fight against fraud to the Union’s financial interests by means of criminal law. Article 12 of this Directive concerns statutes of limitations for investigation, prosecution, trial and judicial decision of criminal offences affecting the Union’s financial interests. What will happen, if a member state decides, after the Directive enters into force in 2019, to extend considerably the limitation periods not only for new but also ongoing investigations, prosecution or judicial proceedings in VAT cases? Under EU law, individuals will not be able to oppose this since in the EU legal order the principle of legality and even legal certainty do not apply to such ‘procedural’ matters as statutes of limitations. Since this has already been settled in the Taricco saga, no ‘legitimate expectations’ of the opposite will arise. Does this mean that EU law has effectively lowered the legal protection of individuals in Italy and any other Member State that considers the principle of legality or legal certainty to be universally applicable, regardless of whether some provisions are labelled as substantive or procedural?

One could allege that I show an exaggerated concern for the rights of serious offenders (or the rights enjoyed by serious offenders at the moment when they should still be presumed innocent, to be precise) and insufficient concern for the rights of citizens adversely affected by the cases of serious VAT fraud (see Ziccardi Capaldo). I would respond that my concern is directed towards the universally applicable standards of legal certainty and fundamental rights. What worries me is that the ECJ provides legal grounds for member states to make potentially serious, sudden and unexpected changes in their criminal procedures and it does this in times when the rule of law in Europe is going through a crisis. What does not reassure me is that this is all for the sake of Union’s financial interests.

Who is the final arbiter of constitutionality in Europe?

Despite the ambiguities in its reasoning, the ECJ should be applauded for correcting its mistake. What we have witnessed in the Taricco saga is that, thanks to convincing legal arguments, a national constitutional court is able to convince the ECJ to make a correction to its previous stance. What definitely ‘helped’ the ECJ to take the ICC’s reservations seriously was the latter’s declaration, made in a non-confrontational yet firm manner, that upholding the Taricco I solution would lead to the disapplication by the ICC of the EU law as interpreted by the ECJ, due to its incompliance with Italian constitutional law.

The ECJ has surely drawn some conclusions from its previous confrontation with a national apex court in Ajos. In this case, the Danish Supreme Court asked the ECJ to reconsider its Mangold jurisprudence relating to the discrimination based on age. The ECJ upheld its previous stance, which lead the Danish Supreme Court to disapply the Mangold doctrine (see more here).

I believe that thanks to the national constitutional doctrines on the ‘conditional’ primacy of EU law (on the ‘conditional supremacy’ of EU law in the UK, see the post by Garner on this blog) as well as to the corresponding EU provisions – the constitutional identity clause in Article 4(2) TEU and the authorisation to apply higher national standards of fundamental rights in Article 53 CFR – national constitutional or apex courts can provide necessary checks and balances on the ECJ enormous judicial power. We rarely consider courts, in particular the apex ones, in terms of their accountability or checks and balances on their power. They are the ones to provide checks and balances on political institutions. But as institutions composed of humans, the ECJ may commit mistakes from time to time. National constitutional or apex courts are well-placed to point out such mistakes. By entering into deliberation via the preliminary ruling procedure, they may convince the ECJ to modify its stance. We should see the occasional tug of war between national constitutional courts and the ECJ regarding the primacy of national constitutional or EU law not as a pathological condition but as a desirable, deliberative mechanisms aimed at balancing judicial power.

Does the recognition of such checks and balance mechanism undermine the primacy of EU law? I believe not as long as the ECJ is able to present its reasoning in a clear and coherent way to justify the solution it has reached to its audience. As regards Taricco II, primacy would have been definitely more seriously undermined, if the ICC had been forced to apply its contrilimiti doctrine. Does this relativize the primacy of EU law vis-à-vis domestic constitutional law (see more in the post by Burchardt)? It should not as the ECJ is anyway obliged to draw from the values recognised within the constitutional systems of Member States under Article 6 TEU. The recognition of the checks and balance mechanism puts higher expectations regarding the persuasiveness of the ECJ’s motives. Admittedly, we have seen in Taricco II that the ECJ has difficulty in drafting a clear, coherent and convincing justification of its rulings. But we should not resign from a necessary checks and balance mechanism fearing that the ECJ is not able to meet the said expectations. The ECJ benefits from the assistance of internal legal services, dealing also with comparative constitutional analysis, thanks to which it should be able to assure that its rulings and their motives address convincingly and preventively potential reservations that different European legal systems may have.

But in such a case, one may ask: ‘who is the final arbiter of constitutionality in Europe?’ as Matthias Kumm once did. In light of the Taricco saga, I hope that in each case the final arbiter will turn out to be the one who presents more convincing arguments.