European Constitutional Identity as the Unamendable Core of the EU Treaties

Abstract

European constitutional identity as an implicit limit to amending EU treaties – General trend towards acknowledging implicit constraints to constitutional amendments as well as empowering supreme courts and constitutional courts to monitor ‘unconstitutional constitutional amendments’ – Substantive limits to revising EU treaties essentially based on Article 2 TEU – Conditionality mechanism rulings as a point of departure for speculating on the emergence of a European constitutional identity – ECJ’s competence to review EU treaty amendments with a self-restrained approach as a last resort mechanism.

Introduction

Legal debates over constitutional identity are long-standing in the field of comparative constitutional law.¹ Constitutional identity is generally conceived as the fundamental basis of a legal order that lies beyond the reach of constitutional amendment power.² The idea of limits on constitutional amendment is anything but new, yet it has gained momentum of late among a variety of scholars and in certain jurisdictions.³ In addition to the eternity clauses many constitutions already contain, more and more jurisdictions across the globe are acknowledging an implicit unamendable core of the constitutional text and, in some cases, courts are even assuming the power to enforce respect for those constitutional boundaries. Because of its contested constitutional foundation,⁴ borrowing the so-called ‘unconstitutional constitutional amendment’ doctrine in EU law appears much more problematic.

Not to be confused with constitutional identity is a similarly worded legal category found in the EU legal system: national identity. While constitutional identity refers to an ad intra dimension – i.e. the intangible nucleus of the constitution – national identity performs an ad extra function as a general principle of EU law governing the relationship between the national legal orders of the member states and the supranational legal system.⁵ In contrast to the national identity clause laid down in Article 4(2) TEU, which has been extensively studied, the notion of constitutional identity as applied to the EU legal order remains a widely neglected legal notion.⁶ Similarly, another nominal clarification is pertinent here. The expressions ‘European identity’ and ‘European constitutional identity’ are not used interchangeably in this research; their meanings differ. Taking the image of two concentric circles as a reference, European identity would be the more general term encompassing European constitutional identity, the latter having a narrower scope of application. Unlike European identity, which alludes to the distinctive legal elements of the EU legal order, I use European constitutional identity more precisely to confine it strictly to the theory of constitutional identity that has been developed by some constitutional courts in the national sphere as a limit on constitutional amendment power.

All the foregoing considerations beg a fundamental question: is there actually such a thing as a European constitutional identity? If so, can it fulfil an ad intra function as an insurmountable limit to treaty revision subject to judicial review?

It is worth giving an example to better illustrate a case where the European constitutional identity is at stake. Let us consider a hypothetical scenario of democratic regression in the EU in which there is also a growing scepticism towards the European project. In particular, there is an initiative on the part of the member states to return to a supranational model in which they have more competences at the expense of the Union. Would the member states be entitled to amend the treaties to abolish the European Parliament? There is little doubt that such a major change would amount to a violation of the European constitutional core. It would be an inadmissible interference with the most elementary notion of the principle of democracy. If this is a somewhat radical example, a more subtle proposal for change could be considered: would the treaties enable the member states to formally maintain the European Parliament but strip it of its co-legislative function? Still, it seems disproportionate to deprive the European Parliament of a competence that is inextricably linked to any parliamentary institution. Finally, to take a more refined example that might genuinely raise a serious doubt, what if the Parliament were simply stripped of the power it shares with the Council to draw up the budget?

However hypothetical and removed from the political agenda these three questions are, they are very interesting from a theoretical point of view. Despite growing discussion about possibly imminent treaty amendments, no one has as yet argued for adding an eternity clause or advanced some other proposal that would empower the ECJ to police ‘unconstitutional constitutional amendments’.⁷

Affirming the emergence of European identity as a legal concept in the EU legal order seems now uncontroversial. Yet it has gone largely unexplored in the scholarship and even more so by the Court of Justice. It has attracted, however, recent attention from the Luxembourg court in two landmark rulings dealing with Regulation 2020/2092, which established a conditionality mechanism to protect the Union’s budget in the context of the EU rule of law crisis.⁸ For the first time, the ECJ explicitly referred to European identity, which has been associated with EU values enshrined in Article 2 TEU. Admittedly, nothing in this judgment connects European identity to EU treaty amendments. Nonetheless, it leaves room to reflect upon the idea of European constitutional identity as a parameter to monitor future EU treaty amendments.

Against this backdrop, the aim of this article is to analyse the original idea of European constitutional identity. To be more precise, my research examines whether this legal notion could be conceived as an implicit limit on EU treaty amendments and, if so, whether the ECJ would have jurisdiction to control its boundaries. For that purpose, this article is structured in three parts. The first part explores what, if any, procedural and substantial limits to EU treaties amendments are operative. Second, the rising idea of European constitutional identity will be analysed in terms of its potential impact on the theoretical framework regulating treaty change. Last, the thorny issue of judicial review of treaty revisions will be taken up.

Amending EU treaties: exploring procedural and substantive constraints

Before starting, a preliminary remark to acknowledge the assumption that the EU can be conceived in constitutional terms and, more importantly, that EU treaties can be deemed, to a considerable extent, as the constitution of the EU.⁹ At this point of European integration, that a process of constitutionalisation is underway is widely recognised; in fact scholars have been studying the EU through the lens of constitutional law for decades.¹⁰ The failed ratification of the European Constitution in 2005 does not refute this assumption, as the Treaty of Lisbon inherited most of that text’s fundamental features. Moreover, the ECJ has consistently endorsed this approach in its case law, even equating the treaties to the ‘basic constitutional charter’¹¹ of the EU.¹² With this endorsement in mind, the purpose of this inquiry is to apply the theory of constitutional change in EU law and, more critically, to examine the idea of constitutional identity at the EU level.

The exclusivity of Article 48 TEU and the autonomy of the EU legal order

Determining the exclusivity of treaty provisions concerning amendments to EU treaties is of utmost importance. Due to the origins of the EU as an entity of international law, it could be argued that the EU treaties might be reformed outside the procedures laid down in Article 48 TEU. If general rules for amending an international treaty were applicable to the EU treaties (Article 39 Vienna Convention on the Law of Treaties), any modification of their content would be permitted as long as unanimous agreement by all parties is reached. Accepting this proposition would entail no procedural or substantial limits to amending the basic texts of EU law, other than the rule requiring unanimity.

Based on two European Coal and Steel Community Treaty (ECSC Treaty) reforms, it was argued that the treaties could be revised without respecting the procedural constraints provided in Article 48 TEU.¹³ According to this view, the lack of constraints on the Member States’ power to amend the treaties, as the masters of the treaties, is implicit in the basic rationale of the Union legal order. However, since the ECSC Treaty was reformed within a transitory period impeding the application of the general clause for amendment (Article 96 ECSC Treaty), its value as a valid precedent is necessarily contested.¹⁴

In contrast to this approach, the ECJ has categorically rejected the possibility of reforming the treaties without regard for the procedures established by the treaties. In very clear terms, the Luxembourg court ruled in the Defrenne case that ‘apart from any specific provision, the Treaty can only be modified by means of the amendment procedure carried out in accordance with the Treaty’.¹⁵ From that point onwards, the ECJ has consistently prohibited various means whereby member states have tried to circumvent the rules for amending the treaties. In this vein, the ECJ has sustained that the member states cannot amend primary law by resorting to joint resolutions,¹⁶ accords concluded with third parties,¹⁷ international agreements¹⁸ or Article 352 TFEU.¹⁹ Additionally, it has insisted that the rules regarding the manner in which the EU institutions arrive at their decisions are not at the disposal of the member states.²⁰ All of these judgments support the autonomous character of the EU legal order – in particular, its separation from the legal framework of internal law on this matter – and strongly assert the exclusive subjection of member states to EU law when reforming the treaties.

These factors allow us to affirm the need for any amendment initiative to comply with Article 48 TEU (or the rules foreseen for the special procedures). Although all amendment procedures share some common features, each one also makes specific stipulations, for instance, concerning the actors entitled to initiate the reform or the majority or other criteria required for ratification by the member states. Significantly, depending on which part of the treaty is to be amended, one procedure or another must be activated. If the ordinary revision procedure could be considered as the default route for amending the treaties, it would be applicable when any of the other procedures are not.²¹ More importantly, a first reading of the treaties seems to indicate that no constraint is imposed when the ordinary procedure is followed. Conversely, when it comes to the simplified procedure, the authors of the treaties included explicit procedural restraints in Article 48(6) TEU: (i) this route for change cannot be used to increase the competences conferred on the Union; and (ii) this route for change is restricted to revising Part Three TFEU relating to internal policies and actions of the EU. Although they are purely procedural in nature and only concern the simplified procedure, they prove that the power to reform the EU treaties is not without limits.

Judicial review of procedural limits: the Pringle case

A distinctive legal feature of the simplified procedure is that the treaty amendment takes the form of a decision within the meaning of Article 288, paragraph 4, TFEU. By contrast, in the ordinary procedure treaty amendment takes the form of an agreement between member states requiring subsequent national ratification. The difference is significant because it decisively affects the possible avenues for judicial review. Unlike decisions adopted under the simplified procedure, whose justiciability has been upheld by the Court since Pringle,²² whether the ECJ is entitled to substantial control over treaty reforms arising out the ordinary procedure remains contested. Before going into a more detailed general analysis of the judicial review of treaty amendments in EU law below, the judicial review of procedural constraints deserves mention here.

In the context of the severe economic and financial crisis that ravaged the EU at the end of the 2000s and in the early 2010s, measures to guarantee the stability of the euro area became imperative. This was the background for European Council Decision 2011/199 amending the treaties via the incorporation of a new paragraph 3 into Article 136 TFEU that provided the legal basis for the creation of the European Stability Mechanism. The amendment was challenged before the ECJ in the Pringle case, which called upon the Court to assess the conformity of the Council Decision with EU law.

A first remarkable feature of the judgment in Pringle is that, although the ECJ lacks competence to assess the validity of primary law, it can review decisions that amend the treaties through the simplified procedure.²³ The ECJ has clarified that it can review the decisions through the preliminary ruling because European Council decisions are acts of the institutions within the meaning of Article 267 TFEU. Thus, the question of whether the Court would be competent to hear a direct action is left unresolved. To affirm its competence, the ECJ relies on its fundamental role as the guarantor of the correct interpretation and application of the treaties under Article 19(1), paragraph 1, TEU. It also delimits the scope of its review with the pronouncement that it will only monitor the procedural conditions laid down in Article 48(6) TEU.²⁴ In the Court’s view, none of those limits were transgressed, confirming the validity of the Decision as in compliance with the procedural requirements set out in the treaties.

Its compliance with one of the requirements, however, warrants reflection, for the decision can only affect Part Three TFEU. For that purpose, the examination of the Court did not restrict itself to the verification of the amendment’s formal relation to that part of the TFEU – which adding a new paragraph to Article 136 TFEU did – but also extended its check to whether the amendment materially affected the rest of primary law.²⁵ In this vein, the Court analysed whether Decision 2011/199 infringed the EU competence in the areas of monetary policy and economic policy coordination, as laid out in Part One of the TFEU. Otherwise, it would be possible for the authorities seeking the amendment to bypass the ordinary revision procedure, whose legal prerequisites are more demanding.²⁶

Even though the ECJ ultimately rejected the argument that the amendment affected primary law outside Part Three, a relevant lesson can be drawn from the Court’s reasoning. As Advocate General Kokott perceptively noted, the Court is not exclusively restricted to procedural review.²⁷ The Advocate General backed up her assertion using two arguments: (i) because the treaties do not forbid substantial judicial review; and (ii) on the basis of an a contrario interpretation of Article 269 TFEU. Since control of the non-affectation condition entails verifying that the amending powers respect primary law outside Part Three TFEU, the distinction between procedural and substantial control is blurred. The legal exercise appears to go beyond mere procedural control. In Kokott’s words, ‘a certain hierarchy of provisions of primary law is created’,²⁸ in that controlling the validity of a decision amending the treaties requires that the parameter of control consist of provisions excluded from the simplified revision procedure.

In sum, the innovative character of this judgment stands out, as the Court, for the first time in its case law, established its competence to exercise judicial review of treaty amendments. Granted, it is circumscribed to judicial control over decisions adopted through the simplified revision procedure. Notwithstanding, it leaves room to wonder whether the Court might subsequently enlarge its powers to include substantive monitoring of treaty amendments approved through the ordinary procedure.²⁹ This is a major issue that will be elaborated further when discussing the jurisdiction of the ECJ.

Substantive limits on amending EU treaties

A fundamental question that must now be addressed is whether, in addition to procedural limits, EU law imposes substantive limits on treaty amendments. As a starting point, neither Article 48 TEU nor any other provision of primary law makes explicit reference to any provision or institution specially protected from modification or derogation. It is clear, then, that no eternity clause was inserted in the treaties.³⁰ At first sight, it seems that the authors of the treaties intended to keep open the possibility for the amendment of any provision. In view of the silence of primary law on this point, an analysis of the Court’s case law seems appropriate, as a number of rulings suggest the opposite conclusion.

The first judicial precedent is Opinion 1/91, which examined the compatibility with the EEC treaties of an international agreement for the creation of the European Economic Area.³¹ The main contention that the Court had to address was the establishment, by means of this international agreement, of a system of judicial control entrusted with interpreting and applying the rules of the international agreement. Specifically, setting up a judicial authority with such powers could conflict with the autonomy of the European Community legal order, particularly with the prominence attributed to the ECJ by the treaties.

In Opinion 1/91, the Court proclaimed that ‘an international agreement providing for a system of courts, including a court with jurisdiction to interpret its provisions, is not in principle incompatible with Community law and may therefore have Article 238 of the EEC Treaty [now Article 217 TFEU] as its legal basis’.³² Following up on that general statement, it declared that ‘Article 238 of the EEC Treaty does not provide any basis for setting up a system of courts which conflicts with Article 164 of the EEC Treaty [Article 19 TEU] and, more generally, with the very foundations of the Community’.³³ In response to a suggestion from the Commission, the ECJ rejected the use of a treaty reform, and more specifically an amendment of Article 238 of the EEC Treaty, to overcome the contradiction between Article 238 and the international agreement.³⁴ The Court therefore distinguished between ‘ordinary primary law’ and the ‘very foundations of the Community’ – among which only the Community judicial system is mentioned.³⁵ According to this distinction, it could be interpreted that the foundations are placed in a hierarchical position above the rest of the primary law, such that no modification or abrogation of the foundations of the Community by means of treaty reform is permitted.

In a case similar to the preceding decision, the ECJ was called upon to adjudicate the compatibility of an international agreement to establish a European Patent Court in Opinion 1/09.³⁶ In contrast to Opinion 1/91, the judicial body established in the agreement would not only have the power to interpret and apply the provisions of the agreement, but also to interpret and apply EU law in that field. Even more problematic was that the agreement deprived national courts from referring a preliminary ruling in the areas covered by the agreement in favour of the European Patent Court. In that regard, the ECJ asserted that the functions conferred to national courts and the ECJ are ‘indispensable to the preservation of the very nature of the law established by the Treaties’.³⁷

A third important case is Opinion 2/13, which can be interpreted along the same lines.³⁸ The ECJ ruled against the accession of the EU to the ECHR, despite the mandate to do so in Article 6(2) TEU. Unlike the previous cases, the contested agreement did not provide for the creation of an ex novo judicial body. Conversely, it proposed granting the European Court of Human Rights competence to hear cases in which the violation of human rights enshrined in the ECHR was committed by the EU. Among the different legal arguments of the ECJ against accession, one of the most decisive ones held that the draft agreement could not guarantee that accession would not affect ‘the specific characteristics of the EU and EU law’.³⁹ The Court went on to state that ‘these characteristics include those relating to the constitutional structure of the EU, which is seen in the principle of conferral of powers …, and in the institutional framework established in Article 13 TEU to 19 TEU’.⁴⁰

In all its previous decisions, the ECJ affirmed that the essential features of the supranational judicial system were part of the basic constitutional structure of the EU. However, these judgments – which all happen to be opinions within the meaning of Article 218 TFEU – appear to recommend a very narrow application of this judicial doctrine. So far, this legal doctrine has placed the emphasis on a very specific trait of the EU institutional system.

Opposing doctrinal approaches

Absent express substantive limits on the treaty amending power, some scholars have interpreted the Court’s rulings as forming a general theory of implicit limits on treaty revision.⁴¹ From an academic standpoint, the debate was initially triggered by the ‘laconic and rather cryptic’⁴² Opinion 1/91 from the early 1990s.⁴³ Still, the discussion can be considered underdeveloped, since it has only been taken up periodically and expressed in very restrictive terms in ECJ case law. Scholars are divided, defending opposing views over the theory that EU treaty amendments are subject to implicit limitations.

According to the theory’s supporters,⁴⁴ a few decisions of the Court provide the justification. Supporters also argue that the theory is consistent with jurisprudence developed by national constitutional courts that defends an untouchable core of the constitution, a strict interpretation of the term ‘amend’,⁴⁵ the theory of supra-constitutionality⁴⁶ and a hierarchy of values within a constitutional text.⁴⁷

On the other side are scholars who consider that the supporters’ interpretation of the Court’s case law is questionable and maintain that member states retain full power to amend the treaties without any substantive constraint.⁴⁸ They submit that the commitment to an ‘ever closer union’ (Article 1 TEU) – as well as the indefinite term of the treaties (Article 53 TEU) – does not prevent member states from reducing the competence of the EU or, for that matter, from transforming the EU in any sense whatsoever.⁴⁹ Since the Treaty of Lisbon entered into force, both mandates must be read in connection to the Article 50 TEU allowing member states to abandon the Union. In this sense, following an a fortiori interpretation, it could be argued that the ability of member states to determine the fate of the EU is unlimited.

It should be clear that the question of implicit substantive limits on EU treaty amendment is highly controversial. Some value may be found by adding to the debate the concept of European constitutional identity as a compelling argument in favour of limiting the member states’ power to modify the treaties without regard for the axiological basis of the European legal order. Such an attempt disregarding Article 2 TEU would not merely represent ordinary revision, but constitutional revolution. Admitting hypothetically unconstrained power of the member states to amend the foundational treaties would imply a procedural or majoritarian notion of democracy; that is, any revision of the treaty would be possible as long as the procedures set out in the treaties are followed. Conversely, I would like to stress that the argument here is that a substantive conception of democracy inextricably linked to shared values– exceptional and restrictively interpreted – poses limits to the member states’ amendment powers.⁵⁰

European constitutional identity: an implicit limit on treaty amendment?

Rediscovering European identity

The idea of European identity is not new in EU law. The legal concept did not appear in the EU’s lexicon for the first time with the twin judgments on the conditionality of European funds in February 2022. On the contrary, these rulings mark the most recent iteration on the long path the idea has taken.⁵¹ The Declaration of the Heads of State and Government on European Identity of 1973 is commonly held as the first document attesting this notion.⁵² Following a common trend in EU law, European identity has undergone a legal transformation from an unwritten principle in a political statement to a general principle of EU law as a result of the Court’s proclamation in the two parallel decisions about the conditionality mechanism.⁵³ European identity was included in some drafts of the treaties, but since the Treaty of Lisbon entered into force, those references are no longer to be found. For the sake of clarity, it should be specified that we speak of ‘European identity’ for the moment in this section because reference is made only to those distinctive elements of the EU’s legal order which make it recognisable. It will not be until later that we refer to ‘European constitutional identity’ in order to incorporate a more concrete approach as the unamendable core of the EU treaties.

With its rulings in the conditionality funds cases, the ECJ marked a new phase of European identity. Above all, both are the first judgments issued by the Court that explicitly and distinctively cite European identity. That said, a precedent of the utmost relevance that cannot be overlooked is the Kadi case.⁵⁴ After a brief comment on this precedent, the two landmark judgments with regard to the conditionality mechanism of EU funds will be examined in more detail. Finally, in connection to the main purpose of this research, the idea of European constitutional identity in the specific context of treaty amendments will be explored.

The key judicial precedent in Kadi

So much has been written on this case that there is no need to summarise it in detail.⁵⁵ A brief reference to the reasoning of the ECJ in the judgment relevant to European identity is sufficient. According to the Luxembourg court:

the obligations imposed by an international agreement cannot have the effect of prejudicing the constitutional principles of the EC Treaty, which include the principle that all Community acts must respect fundamental rights, that respect constituting a condition of their lawfulness which it is for the Court to review in the framework of the complete system of legal remedies established by the Treaty.⁵⁶

On several occasions, the Court has emphasised that the principles of freedom, democracy and, most importantly in this case, fundamental rights constitute ‘the very foundations of the Community legal order’.⁵⁷ Due to the relevance that the ECJ ascribes to them, it asserts that ‘an alleged absolute primacy’⁵⁸ of UN Security Council resolutions is not admissible if they might compromise those foundational principles.

Following this ruling, a large and growing group of scholars has also defended a material conception of European identity.⁵⁹ It is even possible to discern three stages the study of this legal category has gone through. A first wave of academic works on the idea followed in the wake of Opinion 1/91; a second phase began with the Kadi case; and, as will be analysed below, the current phase was initiated with the twin decisions upholding the conditionality mechanism.

Two parallel decisions confirming a long-standing concept

Sitting as a full court, the ECJ issued a much-awaited decision in 2022 in the context of the rule of law crisis in Hungary and Poland.⁶⁰ The two member states had challenged the validity of Regulation 2020/2092 establishing a general conditionality mechanism in the interest of protecting the Union budget. It can be considered the latest mechanism devised to overcome the inefficacy of all the other legal tools meant to deal with democratic backsliding in our continent. In essence, it is based upon the premise that a member state that disregards the principle of the rule of law cannot reliably ensure either sound financial management of EU funds or protection of the Union’s financial interests. Predictably, Hungary and Poland disagreed and filed two actions for its annulment. For the purpose of our inquiry, the most interesting angle involves the development of previous case law about EU values.

With regard to Article 2 TEU, the Court reiterated and reinforced preceding jurisprudence concerning EU values. It recalled that the commitment to EU values was a mandatory, previous criterion for membership in the EU (Article 49 TEU). According to the ECJ, compliance with that prerequisite gives rise to mutual trust between member states that they all share and are committed to respecting those values. Following case law that began with the Reppublika judgment,⁶¹ the ECJ confirmed that Article 2 TEU is a requirement that member states must also continuously fulfil to retain Union membership and the rights from the treaties that come with it.

The reaffirmation of the conditions for EU membership described above paved the way for the crucial paragraph in this pair of judgments. Using identical wording in each decision, the Court stated that:

The values contained in Article 2 TEU have been identified and are shared by the Member States. They define the very identity of the European Union as a common legal order. Thus, the European Union must be able to defend those values, within the limits of its powers as laid down by the Treaties.⁶²

Here, for the first time, the Court utters the words ‘European identity’ in its case law and specifies its composition: the values enshrined in Article 2 TEU. This declaration could be submitted as an endorsement of the line of reasoning advanced by some scholars and a refinement of the preceding ambiguous case law. It renders more precise and at the same time broadens the timid stance adopted in Kadi by widening the scope of European identity in two important ways: first, by no longer limiting the ECJ jurisdiction to instances when fundamental rights are at stake; and second, this pair of judgments enlarge the protection of EU values beyond potential violations stemming from the application of international law. The protection now seems to cover breaches of EU’s constitutional core deriving from national law. Although somewhat boldly, it could be sustained that this judgment emulates the contralimiti doctrine at the supranational level, since the ECJ attributes to itself the task of safeguarding the basic constitutional features of the EU legal order from violations emanating from national law.⁶³ Also worth mentioning is that these two judgments do not solely represent two isolated rulings issued the same day. A year later, again in the framework of the rule of law crisis in Poland, the Court reiterated its association of EU values with the ‘very identity of the European Union’.⁶⁴

At this point, there is little room for doubting the growing application of European identity in EU law. Although it has been directly tied to the EU values in Article 2 TEU,⁶⁵ many arguments have come from the academy enlarging the meaning of that legal notion. In addition to the values in Article 2 TEU, many scholars concur that the constitutional principles of the EU legal order – for instance, principle of primacy, or principle of direct effect – and the four freedoms of the internal market should be included in the identity as a sort of economic constitution of the EU.⁶⁶ Regardless, whichever notion of European identity is subscribed, they all agree on the minimum core: EU values.

In my view, as far as EU values are concerned, a restrictive meaning of European identity should prevail, according to which only the essence of those values must be respected, without rejecting the possibility that other EU constitutional principles can be incorporated into this notion.⁶⁷ Otherwise, the principle would disproportionately restrain the power of member states to design the supranational legal order – as well as their own legal systems – within reasonable boundaries. By way of illustration, there is not a unique constitutional setting faithful to the value of democracy. Different constitutional models may comply with such value and, if they do, they should be accepted. Although it often seems no more than a play on words, European identity poses a limit on national identity, as many scholars noted years ago.⁶⁸ Now, however, it seems that the ECJ is taking this approach to counteract the so-called ‘abuse of national identity’⁶⁹ phenomenon.⁷⁰ In all, the judiciary has been observed to use European identity in terms of three different scopes of application. It could thus be conceived as a triple limit to be invoked independently when international law, national law or, more importantly for our analysis, EU law are applicable.

Much has been said about the meaning and scope of this enigmatic expression of European identity. Doubts persist regarding its potential relevance to the ‘unconstitutional constitutional amendment’ doctrine in EU law, a relatively underexplored field of study. Admittedly, these two judgments issued by the Court confirming the validity of Regulation 2020/2092 are placed in a different context. Nevertheless, when the Court invokes identity, the idea that comes to mind is the judicial doctrine of constitutional identity at the supranational level.

Judicial review: borrowing the doctrine of unconstitutional constitutional amendments?

The doctrine of unconstitutional constitutional amendments: an overview of national experiences

At first glance, the mere expression ‘unconstitutional constitutional amendment’ seems contradictory or even paradoxical.⁷¹ It seems complicit with the similar postulate that a constitutional provision can be unconstitutional. Two main arguments may explain this apparent conundrum. For one, if we accept that the different provisions of the same constitution fall into a hierarchy, when one constitutional norm contradicts another, the one with higher status overrules the other. Second, a distinction must be made between a constitution, which is a valid legal text in force, and a constitutional amendment, which is a later proposal for changing the constitutional text. Keeping this temporal perspective in mind, it comes as little surprise that constitutional amendments sometimes contradict their constitution.

While it has been submitted previously that there are substantive (and procedural) limits on EU treaty amendments, the existence of such limits does not necessarily imply the ECJ’s competence to review them.⁷² Without doubt, the Court’s lack of competence has highly problematic consequences for enforceability, since it leaves the ECJ unable to safeguard substantive limits. The effectiveness of unamendability is therefore closely related to its judicial oversight. From the standpoint of constitutional design, this argument strongly favours the monitoring of treaty revisions. Renouncing the ECJ’s jurisdiction would also pose serious problems from the perspective of the principle of the rule of law, inasmuch as the political power to change the constitutional basis of the European polity would not be subject to judicial scrutiny. The Pringle case settled the ECJ’s competence to ensure the compliance with the procedural requirements set out in Article 48 TEU under the simplified procedure. Whether the Court could exercise this control in case of breach of substantive limits in the framework of the ordinary procedure remains unanswered, as do the questions of which legal avenues provide access to the Court and what standard of review would be applied.

A brief comparative glance reveals that there are different models concerning judicial review of constitutional amendments. As a starting point, constitutions rarely grant a constitutional court the competence to review constitutional amendments.⁷³ In the EU, the German Federal Constitutional Court has asserted its competence to review constitutional amendments since the seminal Klass case⁷⁴ in 1970.⁷⁵ Although it is true that the Bundesverfassungsgericht has exercised this prerogative on several occasions, even in the absence of an express authorisation in the Basic Law, it has never declared a constitutional amendment invalid for conflicting with the eternity clause. The French case is somehow different because, even if the constitution contains an eternity clause shielding the republican form of government, the Conseil Constitutionnel has declined control over constitutional amendments.⁷⁶ The Italian example is even more interesting. Not only does the Corte costituzionale enjoy competence to review constitutional amendments, it has also assumed the jurisdiction over implicit limits outside the eternity clause.⁷⁷ This case is particularly relevant because in it an authoritative judicial body declared its competence to review a constitutional amendment affecting basic constitutional features not protected by an explicit provision. Judicial review of constitutional amendments – whether the legal system explicitly or implicitly foresees the limits – is clearly not solely a theoretical hypothesis but rather a practice existing in various jurisdictions in the EU (and also beyond our continent).

The Court’s competence to review substantial limits to treaty amendments

The question of whether the ECJ has jurisdiction to hear cases challenging EU treaty amendments has no straightforward answer. Since the Treaty of Lisbon, which significantly expanded the ECJ’s purview, the Court has jurisdiction over all matters concerning EU law with very limited exceptions, especially in the field of Common Foreign and Security Policy.⁷⁸ As a rule, then, the ECJ is granted authority to adjudicate and only exceptionally is such authority to be denied. Accordingly, any restrictions imposed on the Court’s jurisdiction must be expressly justified and narrowly interpreted in order to guarantee effective judicial protection. Admittedly, no provision explicitly assigns the responsibility for such review to the ECJ, but, importantly, the treaty no longer contains a specific norm stipulating in which areas the Court may intervene.⁷⁹ Currently, the treaties do not provide any restriction on the jurisdiction of the Court to review treaty amendments. Additionally, scholars have argued that, since Article 269 TFEU ‘explicitly lays down a restriction in other circumstances, it can be argued, a contrario, that the Court has the competence to review both formal and substantive aspects of treaty amendments’.⁸⁰

Careful examination of the ECJ case law may also invite greater caution. The Court differentiates, for the purposes of judicial review, between treaty amendments taking the form of treaties between Member State and acts of the EU institutions.⁸¹ Since the jurisdiction of the Court over acts of EU institutions has already been analysed in detail, it is now time to consider the justiciability of amendments passed by treaty.

The ECJ has declined to settle the validity of primary law in its case law.⁸² A point of reference is the LAISA case, concerning a motion for annulment action and a claim for damages against the annex of the accession treaty of Spain and Portugal and the adjustments to the treaties resulting directly from the act of accession. The ECJ held that the contested provisions do not constitute an act of the Council within the meaning of Article 173 of the EEC Treaty [Article 263 TFEU], but rather provisions of primary law. For that reason, the Court recused itself from considering the legality of the provisions so, by consequence, the actions for annulment were declared inadmissible. With regard to the action for damages, the applicants’ claims were also declared inadmissible using similar arguments.⁸³ Since then, this line of reasoning has become a recurrent formula of the ECJ when accession treaties are in question.⁸⁴

As for the procedure for ordinary revision, which is our main concern here, the ECJ has also shied from assuming its competence to rule on the non-contractual liability of the Community. It has considered that the treaties – whether in the original version or including any amendment – ‘constitutes neither an act of the institutions nor an act of the servants of the Community. The Treaties cannot, therefore, give rise to non-contractual liability on the part of the Community.’⁸⁵

Other legal actions and standard of review

It follows from the Court’s case law that treaty amendments are not subject to ECJ review if they are challenged by an action for annulment or action for damages. However, those legal actions do not exhaust all the legal remedies for control of substantive limits before the Court. Treaty amendments might still be reviewable via an infringement action or a preliminary ruling. As for the former, some scholars have already argued that proceedings against all member states adopting an amendment by common accord within the meaning of Article 48(4) TEU could be initiated on the basis of Article 258 TFEU.⁸⁶ Similarly, seeking preliminary reference on the validity of treaty amendments seems unproblematic. Admittedly, according to Article 267 TFU, the treaties are not legal acts reviewable by means of this indirect action when the validity of Union acts is being assessed. Nevertheless, as was explained above, treaty amendments only become primary law when they are in conformity with the legal text of the treaties in force. Moreover, the preliminary reference has been designed and interpreted by the Court very flexibly, such that it is only declared inadmissible under limited and strict circumstances.

To state that the Court has competence to rule on the validity of treaty amendments is quite different from stipulating the standard of review to be used by the Court of Justice. As a general consideration, the high threshold of legitimacy required through the ordinary revision procedure, given the requirement of unanimity vote of member states, warrants deference from the Court. For the same reason, presuming the validity of treaty amendments would avoid an undesirable effect in terms of institutional balance.⁸⁷ Otherwise, the Court would be taking on a prominent function – somehow comparable to a negative treaty-making power – which political actors are primarily expected to fulfil. In addition to the possible separation of powers issue, the severe legal consequences attached to a declaration of annulment make other less far-reaching remedies advisable.⁸⁸ The Court should therefore first attempt to accommodate to the limit of reason the contested amendment within the treaties.

Last but not least, it is necessary to determine the level of impact on European constitutional identity considered admissible. In other words, not every contact with the untouchable core necessarily entails violation. Roznai differentiates three standards of review: minimal effect; disproportionate violation; and fundamental abandonment.⁸⁹ In his view, cases where constitutional amendments carry a high degree of legitimacy, as happens to be the case when the ordinary revision procedure is followed at the EU level, warrant application of the fundamental abandonment yardstick. That is precisely the standard imposed by the German Constitutional Court, by which only extraordinary infringements of unamendable principles could justify judicial invalidation.

Setting such a high threshold of control, so that the ECJ can only intervene in the most serious violations, undoubtedly provides weak judicial protection for non-amendable principles. In any case, in my view, this is the most appropriate position to strike the right balance in order to take into account also the implications of a stronger role for the Luxembourg Court from a normative and institutional perspective.⁹⁰ It allows preservation of the basic principles of a legal system without establishing an excessively rigid theory of constitutional change and, at the same time, it balances more delicately the relation between the holder of the constitutional amendment power and the courts. In the end, this judicial doctrine ought to be used as an ultima ratio instrument only to respond to the most serious attacks on the essence of EU values and should be undertaken with utmost precaution. Finally, it should also be seen as an additional mechanism for safeguarding the European constitutional foundations, so that the effectiveness of their protection is reinforced as a result of the joint application of the different instruments existing in the European legal order.⁹¹

Conclusions

European identity is emerging as a material legal category in EU law. The consensus that EU values are an integral part of that notion is broad. According to some prominent scholars, its meaning could comprehend some of the essential characteristics of the European legal order. Notwithstanding, here we propose a restrictive approach whereby only the essence of values enshrined in Article 2 TEU must be respected, without rejecting the possibility that other EU constitutional principles can be incorporated into this notion. In following this attitude, a middle ground between respecting the fundamental axiological basis of the EU and according member states enough autonomy to determine the constitutional framework of the EU (and that of their national legal orders) can be reached.

Determining the scope of application of this enigmatic notion has proven much more problematic. Conceiving European identity as a limit to EU treaty amendments – or, put differently, to advocate for the idea of constitutional identity at the supranational level – is highly controversial. That said, the two decisions handed down by the ECJ about the conditionality mechanism leave enough room for speculation about possible implications. It is must also be underscored that the two judgments are not isolated instances without continuation in the case law, as the same proclamation was upheld by the Court a year later.

Besides procedural limits to EU treaty amendments, it has been submitted that implicit substantive limits to the EU treaty-making power also exist. On this point, jurisdictions throughout the world, including in Europe, have deduced implicit limits on constitutional amendments in what seems to have become a popular trend in comparative constitutionalism. Still, even at the expense of undermining the effectiveness of unamendable principles, the mere existence of limits to amendment does not imply judicial control. After Pringle, there can be no doubt of the ECJ’s authority to review procedural limitations in the framework of the simplified revision procedure. As for revisions made through ordinary procedure, however, nothing in the twin decisions issued by the Court supports empowering the ECJ to exercise review. Both implicit limits to treaty amendments as well as judicial review in this field are highly contested issues in EU law. Here a theoretical account has been proposed in favour of the recognition of such limits and the Court of Justice’s power to monitor them subject to restriction.

Only time will tell whether European identity is merely a rhetorical device in the hands of the Luxembourg court or enforceable as a sort of European counter-limit in cases of serious threat to the basic principles of the EU legal system. Thus far, two years after its proclamation, it has never been applied. Adopting this bold approach would, in any event, represent an evident step forward in the constitutionalisation of the EU.