Introduction
In its judgment no. 62/2016, the Belgian Constitutional Court, widely regarded as one of the most Europhile constitutional courts in the EU, has unexpectedly added itself to the list of constitutional courts which do not fully accept the European Court of Justice’s point of view concerning the primacy of EU law over national constitutions.Footnote 1 In the present case note, we aim to contextualise this judgment and to indicate its consequences for the full effect of EU law within the Belgian legal order. This judgment seems to follow the Bundesverfassungsgericht’s example by setting some limits to the primacy of EU law over the Constitution. As the limit related to the Belgian national identity was articulated more explicitly than the one related to ultra vires review, and as it is not clear whether the limit concerning the basic values of constitutional rights protection can stand on its own or whether it is a part of identity review, we will mainly focus on the former limit.
In the first section, we depict the judgment itself, which stands out both because it denied standing to a wide range of petitioners, and because it developed criteria limiting the primacy of EU law over the Belgian Constitution. As the judgment itself is very succinct on this important novelty, the subsequent sections aim to provide some context and suggest answers to several questions left open by the Court. The second section explains the Constitutional Court’s case law applying EU law, indicating that the Court understands its role very well in a context of multilevel constitutionalism, by indirectly reviewing legislation against primary and secondary EU law, by engaging in a very intensive preliminary dialogue with the Luxembourg Court and by applying all the procedural obligations and limits that follow from the principle of full effect of EU law. As judgment no. 62/2016 can be seen as a departure from the Constitutional Court’s usual stance towards EU law, this section also tries to provide an explanation for the Court’s decision. In the third section, we place the judgment in its broader context regarding the hierarchy between treaty law, EU law and the Constitution, indicating that, in principle, EU law still trumps constitutional provisions, while noting that this hierarchy is determined by the Constitution itself. In the fourth section, we discuss the most important unanswered questions raised by the annotated judgment: which procedure applies to the Constitutional Court’s raising an ultra vires or a national identity issue? Which Court would be competent to conduct the review? Which criteria would have to be used when addressing such an issue? Could the Constitutional Court declare a piece of EU law unconstitutional and what would the consequences of such a statement be? And, is it even possible to define what constitutes Belgian national identity? We conclude by nuancing the judgment’s consequences for the full effect and uniform application of EU law in Belgium.
Judgment no. 62/2016
A lack of standing for challenging the legislative approval of the TSCG
In the annotated judgment, the Constitutional Court rejected several actions for annulment filed against the Act approving the Treaty on Stability, Coordination and Governance in the Economic and Monetary Union (TSCG),Footnote 2 the cooperation agreement between the Federal State and the Federated entities for the execution of Article 3.1 TSCG, and the Flemish Decree approving that cooperation agreement.
The actions for annulment were filed by various non-profit organisations and individual applicants, covering the entire political spectrum and a wide range of societal interests. The Ligue des droits de l’homme (Human Rights League) claimed standing in this case because no human right, and especially no social right, would be safe under a scheme of relentless austerity. Several trade unions and their leaders and representativesFootnote 3 claimed standing because a unilateral focus on austerity would undermine the collective negotiations between the representatives of employers and employees, and because it would threaten the automatic indexation of wages in Belgium.
Several individual applicants argued that the TSCG would impair their ability to take part in the political process, as it would divert significant powers away from the national parliaments.Footnote 4 Others alleged that the universities, governmental agencies and non-profit organisations they work for would receive less public funding or even lose all their support because of austerity measures. Some blamed the loss of their jobs in the public or private sector on budget cuts. One member of the Parliament of the German-speaking community cited his duties as an MP, which included deciding upon budgetary matters. One student alleged that the austerity measures would lead to increased tuition fees. Most individual applicants also invoked their citizenship as sufficient ground for standing in this case.
Natural and legal persons are allowed to initiate an action for annulment before the Constitutional Court, provided that they have an ‘interest’ (belang, intérêt) in the outcome of the case.Footnote 5 The Court does not accept an actio popularis, i.e. persons taking up the defence of the constitutionality of legislation without having any personal interest at stake.Footnote 6 Natural persons only have an ‘interest’ if their situation could be directly and adversely affected by the piece of legislation they challenge. Non-profit organisations defending their societal goals before the Court must additionally prove that the goals mentioned in their bylaws differ from their members’ individual stakes, amounting to a collective interest,Footnote 7 which must nevertheless be distinguishable from the common good. Individual MPs, when not invoking their own personal interests, must substantiate their functional interest by showing that the legislative provisions they challenge affect ‘the prerogatives of the individual exercise of their mandate’.Footnote 8 As trade unions lack legal personality, they may only challenge legislation which impairs their involvement in the functioning of public services and only in matters for which the law recognises them as separate entities.Footnote 9
Commentators agree that the Constitutional Court maintains a low threshold when applying these criteria, and generally accepts the standing of all applicant categories without a qualm.Footnote 10 In the annotated judgment, however, the Constitutional Court denied standing to all applicants, and hence declared all actions for annulment inadmissible.
The trade unions could not challenge the legislation approving the TSCG, because this treaty does not directly affect their functioning.Footnote 11 Likewise, the MP in the Parliament of the German-speaking community was not able to prove that the TSCG affects the prerogatives of the individual exercise of his mandate,Footnote 12 nor were the citizens able to prove that their right to vote would be directly and adversely affected.Footnote 13
Concerning the citizens’ interest in budgetary matters, the Court refers to the Legislature’s prerogative to decide on annual income and expenditure.Footnote 14 Parliament is granted a wide margin of discretion in deciding on the socio-economic choices to be made when determining the annual budget. Falling under this discretion is Parliament’s right to set medium-term budgetary goals in collaboration with other institutions, and to consent to international treaties concerning these goals. This discretion applies a fortiori if such a treaty is concluded between countries sharing the same currency, especially if they have mutually undertaken prior international obligations concerning joint economic policy and budgetary restraint.Footnote 15 Moreover, Article 174 of the Constitution, the provision anchoring the Parliament’s prerogatives in budgetary matters, is not violated, since the TSCG does not impair the national parliaments’ discretion in implementing measures for respecting the budgetary constraints.Footnote 16
As far as the duty to uphold the social rights enshrined in Article 23 of the Constitution is concerned, the Court notes that the TSCG’s purely mathematical requirements, and the budgetary framework it creates, leave the substantive policy choices of the different levels of government and their parliamentary assemblies unaffected. Mere consent to these budgetary treaty provisions does not directly and adversely affect the applicants’ individual situation. Only specific implementing measures aimed at attaining budgetary goals might have such an effect.Footnote 17
The Court thus concludes that none of the actions for annulment are admissible. The TSCG’s effect on the applicants’ situation is too indirect to grant them standing. Clearly, the line of reasoning adopted by the Constitutional Court is transposable to just about any potential applicant, which makes it virtually impossible to challenge legislation approving treaties on budgetary matters before the Constitutional Court. As this decision deviates from the Constitutional Court’s usual approach concerning the admissibility of actions for annulment, one might conclude that the Constitutional Court did not want to further complicate a significant step in the process of European integration. One might also conclude that the Court wanted to strike a balance between this part of the judgment and the subsequent part, which could be interpreted as a potential – albeit rather hypotheticalFootnote 18 – hurdle to further European integration.
An obiter dictum concerning Belgium’s national identity
The aforementioned reasoning sufficed to deny standing to the petitioners. Nevertheless, the Court added that the TSCG not only contains budgetary rules, but also attributes new competences to the Commission and to the Luxembourg Court. It then developed a new standard for examining the constitutionality of such rules:
‘When approving a treaty which [attributes new competences to EU institutions], the legislature must respect Article 34 of the Constitution. By virtue of that provision, the exercising of specific powers can be assigned by a treaty or by a law to institutions of public international law. While these institutions may subsequently decide autonomously about how they exercise these competences, Article 34 of the Constitution cannot be interpreted as granting an unlimited licence to the legislature, when approving that treaty, or to the said institutions, when exercising their attributed powers. Article 34 of the Constitution does not allow a discriminating derogation to the national identity inherent in the fundamental structures, political and constitutional, or to the basic values of the protection offered by the Constitution to all legal subjects’.Footnote 19
This is indeed an obiter dictum, since the Court admits, in the following paragraph, that the TSCG does not attribute new competences to the Commission or to the Luxembourg Court after all: it states that the Commission can only make non-binding suggestions, without limiting the National Parliaments’ prerogatives, and that the Luxembourg Court was not granted competence to review the Member States’ budgetary policies.Footnote 20 Developing the aforementioned review standard was therefore not necessary for rejecting the actions for annulment.
By mentioning this obiter dictum, the Constitutional Court seems to acknowledgeFootnote 21 its jurisdiction to review both the legislative provisions approving treaties which attribute new powers to the EU and the legislative provisions transposing or implementing secondary EU law against three constitutional limits which it had never mentioned previously. It had already made clear that such legislative acts fall under its review, but this case law had only concerned a review against its usual reference provisions. Moreover, the Court had always applied a significant degree of caution in this regard, and has never found any violation of the Constitution when conducting such a review.Footnote 22 The development of these new limits is what makes this judgment a novelty.
The two limits explicitly mentioned seem to reflect the Bundesverfassungsgericht’s Lisbon- and Honeywell-jurisprudence. In this case law, the German Federal Constitutional Court grants itself the power to conduct a constitutional identity review of legislation transposing or implementing EU law, which encompasses a constitutional rights review.Footnote 23 The other Lisbon- and Honeywell limit, ultra vires review, is formulated less explicitly in the annotated judgment: it can only be read in the statement that the EU organs may not use the attributed powers as an ‘unlimited licence (…) when exercising their attributed powers’. Moreover, ultra vires review is inherently present in the logic of a limited attribution of powers based on a constitutional provision. The Court probably refrained from further elaborating upon this limit because the case at hand did not involve a norm of secondary EU law.
The Court’s new review standard not only reflects the Bundesverfassungsgericht’s jurisprudence, it also draws upon the text of Article 4(2) TEU when defining the notion of national identity. From a comparative point of view, this is an atypical choice of words, as the other constitutional courts which had already adopted a controlimiti doctrine, founded their national identity rhetoric on their own constitutions. The Belgian Constitutional Court’s phrasing should, however, not be regarded as ‘constitutionalising’ Article 4(2) TEU. If the Constitutional Court conducted a national identity review, it would still examine a piece of legislation against a specific constitutional provision, which it would then describe as being part of the Belgian national identity. It cannot review legislative acts against Article 4(2) TEU. It probably used the language of Article 4(2) TEU because the Belgian Constitution does not provide the means to define this concept.Footnote 24 Another explanation for the Court’s choice of words is that it uses the text of Article 4(2) TEU in order to indicate that it is involved in an indirect dialogue with the Luxembourg Court.Footnote 25
The Constitutional Court does not explain what it means by the third limit, ‘the basic values of the protection offered by the Constitution to all legal subjects’ and it does not clarify whether this is an independent limit or an integral part of Belgian national identity. At least some of the human rights provisions in the Constitution can be deemed part of Belgian national identity,Footnote 26 especially insofar as they offer more extensive human rights protection.Footnote 27 If the German example is to be followed, national identity potentially encompasses the core of all constitutional rights.Footnote 28 Reading this third limit as referring to substantive human rights review would add little to the second one, but reading it as a procedural safeguard does make sense. Article 159 of the Constitution allows all judges to review executive acts against all higher norms, whereas article 142 guarantees individual access to the constitutional review of legislation. The third limit might be read as safeguarding this full access to judicial review, even if EU law dictates otherwise. It might therefore be read as criticising the Luxembourg Court’s Melloni judgment,Footnote 29 but, even in this interpretation, the third limit is linked to Belgian national identity, as Article 159 of the Constitution was a reaction to the Dutch King Willem’s policies.Footnote 30
Conclusion
Judgment no. 62/2016 by the Belgian Constitutional Court annulled neither legislative approval nor the further application of the TSCG in Belgian law. Its atypical adjudication of the petitioners’ lack of standing indicates that it cherishes no ambitions to jeopardise EU integration. Nevertheless, it does contain a warning for Belgian legislators and EU organs alike: Article 34 of the Constitution forbids attributions of powers to the EU, and the application thereof by the EU organs, insofar as they encroach upon Belgian national identity or the basic values of constitutional rights protection. More implicitly, it also disallows ultra vires decisions by EU organs. In adopting this approach, the Constitutional Court has made the hierarchy between treaty law, EU law and the Constitution even more complex than it already was. The Belgian view on this hierarchy will therefore be discussed in the third section, below. Before doing so, however, we will try to explain why the annotated judgment makes it an outlier in the Constitutional Court’s case law concerning EU law.
Judgment no. 62/2016 within the Belgian Constitutional Court’s case law on EU law
Although judgment no. 62/2016 is in line with the well-established case law of several other constitutional courts,Footnote 31 it may still come as a surprise, since it deviates from the Belgian Constitutional Court’s case law regarding EU law. Commentators agree that the Belgian Constitutional Court has always adopted a very Europhile stance.Footnote 32 EU law indeed plays an important role in the Court’s jurisprudence, both from a substantive and from a procedural point of view. We will first describe this role before explaining why the annotated judgment took a different path.
Indirect review of legislation against EU law
First, the Constitutional Court granted itself jurisdiction to review legislation against EU law. Article 142 of the Constitution and the Special Majority Act of 6 January 1989 on the Constitutional Court only grant it the competence to review legislation against certain constitutional provisions and against some infraconstitutional provisions regarding the division of legislative competences between the federal State and federated entities. Nevertheless, since 1989, the Court has involved, along with other high norms, the principle of equality and non-discrimination laid down in Articles 10 and 11 of the Constitution in its review procedure. Indirect reference provisions of EU law may be laid down in all provisions of primary or secondary EU law.Footnote 33 Since 2004, it has also combined the constitutional rights provisions, which have become direct reference norms, with international and European treaties containing analogous human rights provisions.Footnote 34 This enables the Constitutional Court to apply the Strasbourg and Luxembourg Courts’ case law in its own human rights review and to update the Constitution’s human rights catalogue, which largely dates back to 1831, thus avoiding conflicts between constitutional review and supranational review. Furthermore, this technique ensures the most extensive human rights protection by requiring that limitations to human rights meet both the criteria laid down in the Constitution and the ones laid down in the treaties.Footnote 35
As a consequence of these techniques, EU law is applied in about 10% of the Court’s judgments and the Luxembourg Court’s case law is mentioned or cited in about 15% of all judgments.Footnote 36 References to the Convention and the Strasbourg Court’s case law account for another 20% of all judgments.
Procedural application of EU law
Next to the aforementioned substantive review, the Constitutional Court also applies the Luxembourg Court’s case law which imposes procedural obligations or limitations on national judges.Footnote 37 It guarantees the full effect of EU lawFootnote 38 by respecting, for example, the obligation of consistent interpretation,Footnote 39 the duty to take provisional measures,Footnote 40 the duty to raise points of EU law of its own motion,Footnote 41 and the prohibition on maintaining the effects of annulled legal provisions.Footnote 42 The Court also grants standing to the Commission whenever EU law requires so.Footnote 43
Furthermore, the Belgian Constitutional Court maintains a very intensive preliminary dialogue with the Luxembourg Court, having referred, until now, 90 questions to the ECJ in 26 referring judgments.Footnote 44 The questions raised concern both the interpretation and the validity of EU law. Generally,Footnote 45 the Court respects the duty to refer laid down in Article 267(3) TEU, and usually implements the Luxembourg Court’s answers correctly.Footnote 46
Concerning human rights review, the Constitutional Court has also referred to the Melki judgment, which sets some limits on the procedural priority of centralised human rights review within the scope of EU law.Footnote 47 It has even requested legislative amendments to the Special Majority Act on the Constitutional Court in order to remove any doubt about the Belgian priority rules’ compatibility with the Melki judgment.Footnote 48
Europhilia with limits?
The aforementioned analysis shows that the Constitutional Court systematically aims to respect the principles of full effect and uniform application of EU law. Judgment no. 62/2016 seems to be inconsistent with that strategy, insofar as it subjects the primacy of EU law over the Constitution to some limitations. This calls for an explanation.
Komarek brilliantly described how constitutional courts are the main victims of the Luxembourg Court’s rigorous application of the principles of primacy, full effect and uniform application, because this case law renders them unable to use the full ambit of their own procedural potential within the scope of EU law, and because it ‘displaces’ them to the benefit of the ordinary and administrative courts, which are sometimes even allowed to set constitutional jurisprudence aside.Footnote 49 The Luxembourg Court’s emphasis on the full effect and uniform application of EU law thus affects the full effect and uniform application of national constitutions. Nevertheless, despite the ECJ’s vision, both ‘full effects’ are of equal importance and neither the Luxembourg Court nor the national constitutional courts should undermine each other’s effectiveness.
The Melloni judgment constitutes a new step in downsizing national constitutional review within the scope of EU law,Footnote 50 as it adds a substantive impediment to the aforementioned (see nn. 38-46) procedural types of displacement. On the one hand, it renders constitutional review impossible within the scope of EU law insofar as the Constitution offers a more extensive human rights protection than the Charter, and on the other, it jeopardises the uniform application of the Constitution by giving rise to distinct review standards within and outside of the scope of EU law.Footnote 51
The Belgian Constitutional Court, which refers to Luxembourg and Strasbourg case law on a regular basis, has not yet mentioned the Melloni judgment. Arguably, this indicates that Melloni has caused some discontent among the constitutional judges, who, in turn, felt obliged to impose some countermeasures of their own. This statement is supported by the substantive potential of national identity review, which can be used to overrule the Melloni line of reasoning.Footnote 52
Judgment no. 62/2016 and the hierarchy between EU law and the Constitution
In the annotated judgment, the Constitutional Court makes three statements about the hierarchy between EU law and the Constitution. First, the hierarchy is not governed by EU law, but rather by Article 34 of the Constitution. Second, Article 34 of the Constitution ranks secondary EU law higher than the Constitution. Third, this primacy is limited by the three aforementioned standards. These statements logically imply that the Constitution is ultimately the highest norm, as it can determine the hierarchy between the highest norms either way. The Court made no mention of Article 33 of the Constitution, although it is inextricably linked to Article 34. Article 33 stipulates that all powers emanate from the Nation and that they must be exercised as determined by the Constitution. Article 34 is both an exception to, and an application of, this general rule allowing the legislator to attribute the exercise of specific powers to international organs. The latter provision was written into the Constitution in 1970 to provide an ex post constitutional basis for Belgium’s adherence, after the Second World War, to several international organisations able to make decisions binding the organs mentioned in the Constitution.Footnote 53
Judgment no. 62/2016 adds additional complexity to the discussion concerning the hierarchy between international law, EU law and the Belgian Constitution. This question had already been extensively debated both by the highest courts, and in legal doctrine.Footnote 54
A clear majority of legal doctrine expresses the opinion that the Constitution outranks international law, referring to Article 33 of the Constitution and the notion of constitutionalism itself, adding that the legislator may not violate the Constitution by approving unconstitutional treaties.Footnote 55 Some authors also note that the TEU, the TFEU and the Charter should be considered, in this regard, to themselves be norms of plain international law.Footnote 56 Concerning the connection between the Constitution and secondary EU law, most authors refer to Article 34 of the Constitution, in which they read constitutional permission allowing secondary EU law to deviate from the Constitution.Footnote 57
The Cour de cassation ranks all international law possessing direct effect higher than all domestic law, including the Constitution.Footnote 58 It argues that the primacy of international law ‘follows from the nature itself of international law’ and constitutes an unwritten general principle of law. Concerning the hierarchy between the Constitution and secondary EU law possessing direct effect, the Cour de cassation simply grounds the latter’s primacy on the Luxembourg Court’s case law.Footnote 59
The Council of State’s legislative division ranks the Constitution higher than international law, as it systematically advises that an unconstitutional treaty may not be ratified before the Constitution has been amended.Footnote 60 Concerning the hierarchy between the Constitution and secondary EU law, both the legislative and the administrative division of the Council of State refer to Article 34 of the Constitution, stating that once the right to exercise a power has been transferred to an international organisation, that organisation may exercise it without being bound by the Belgian Constitution.Footnote 61
The Constitutional Court’s case law also distinguishes the hierarchy between the Constitution and international law from the hierarchy between the Constitution and secondary EU law. It derives the primacy of the Constitution over international law from article 167 of the Constitution, which stipulates that treaties ‘take effect only after they have received the approval of the [competent legislator]’. Such an approval Act is subject to constitutional review as is any other Act of Parliament but, as there is only one provision, stipulating that the treaty shall have full effect, the Court’s review must cover the treaty’s content. Adding an additional argument, the Constitutional Court states that the legislator should not do indirectly, by ratifying a treaty, what may not be done directly, i.e. violate the Constitution.Footnote 62 In these judgments, the Court adds that it will review approval acts with considerable self-restraint because they are not unilateral acts of sovereignty and do not bear legal consequences outside the Belgian legal order. The annotated judgment contains the same remark.Footnote 63 Because of this self-restraint, the Constitutional Court has never declared an Act approving a treaty unconstitutional.
The Constitutional Court’s position concerning the hierarchy between the Constitution and secondary EU law has long remained unclear. As a basic tenet, the Court declares itself competent to review, against the Constitution, any piece of legislation which transposes or implements an act of secondary EU law, but it adds that it must nevertheless take this circumstance into account when conducting its review.Footnote 64 In practice, it has only declared such legislation unconstitutional insofar as the Directive it transposed granted the Member States a significant degree of discretion as to the means employed in order to reach the Directive’s goals.Footnote 65 If the act of secondary EU law does not grant the Member States any discretion, the Court refers the case to the Luxembourg Court for a question on the act of secondary EU law’s validity.Footnote 66
The first mention of Article 34 of the Constitution in this regard occurred in the CREG-judgment. The Second and Third Electricity Directives oblige national legislators to confer vast regulatory powers upon their independent energy regulatory offices (in Belgium: the CREG). The implementing legal provisions were challenged due to a lack of accountability towards the Minister and Parliament. In its proportionality test, the Constitutional Court first stated that the preamble to the Third Electricity Directive allows the existing parliamentary and judicial monitoring of the independent energy regulators within each Member State.Footnote 67 It also noted that the Council of State’s normal review of executive acts applies to decisions taken by the CREG, and that the federal Parliament approves its budget and monitors the Minister who receives annual reports from the CREG.Footnote 68 The Court then added: ‘Insofar as the previous considerations would not suffice to justify why persons subjected to the CREG’s decisions do not enjoy the guarantee that this decision is taken by an administration which is part of the Executive, this is justified, by virtue of Article 34 of the Constitution, by the requirements following from EU law’.Footnote 69
This judgment in no way established the absolute primacy of secondary EU law over the Constitution. Article 34 of the Constitution was used as the sole argument in a proportionality test. The Court thus read the provision as granting the legislator a very large, yet not unlimited, margin of appreciation when transposing the obligations which follow from secondary EU law. The judgment, therefore, did not preclude the development of Belgian Lisbon and Honeywell jurisprudence. In any event, it already seemed to indicate that the primacy of EU law over the Constitution is based on the Constitution itself.
Judgment no. 62/2016 clarifies this point of view and specifies the limits to the primacy of EU law over the Constitution. By introducing an identity review and an ultra vires review, the Court shows that it will not unconditionally accept the primacy of secondary EU law over the Constitution. It does, however, conditionally accept this primacy, insofar as the EU’s decision does not go beyond the competences attributed to the EU, and insofar as it does not breach Belgium’s national identity or the basic values of constitutional rights protection. This judgment should therefore not be read as a radical reversal of the Constitutional Court’s Europhile strategy, nor as an overall rejection of the primacy of EU law, although it is surely a rejection of the absolute primacy of EU law, as laid down in the Luxembourg Court’s case law.
It is true that the national identity limit identified in judgment no. 62/2016 could be interpreted as a threat to EU integration. The likelihood that this threat would be put into practice is, however, small. The limits drawn by the Constitutional Court clearly resemble the Bundesverfassungsgericht’s limits. However, given the Constitutional Court’s overall Europhile attitude,Footnote 70 it will probably apply these limits with even more restraint than the Bundesverfassungsgericht does.
Moreover, the constitutional provisions at play differ significantly. The Belgian Constitution possesses no specific provision on EU integration, whereas the German Constitution contains an Europarechtsartikel in Article 23 of the Grundgesetz. That provision specifies the limits of EU integration in a much clearer way than Article 34 of the Belgian Constitution does. Article 23 of the Grundgesetz also refers to the absolute limits of EU integration laid down in Article 79(3) of the Grundgesetz (the Ewigkeitsklausel). By contrast, Article 34 of the Belgian Constitution does not place any substantive limit on the transfer of powers to institutions of public international law, nor does it contain an Ewigkeitsklausel.
In the next section, we shall discuss the procedure the Constitutional Court should follow when operationalising its review, the criteria it should use, and the content of the Belgian national identity. These features will reduce the likelihood of conflict even further.
Questions not answered by Judgment no. 62/2016
Another significant difference between the German example and its subsequent Belgian adoption is the latter’s brevity. Whereas the Bundesverfassungsgericht developed a detailed and nuanced theory of limitation and dialectic over a period of four decades and with several elaborate judgments, the Belgian Constitutional Court established the same doctrine seemingly overnight in a short obiter dictum, obviously without the same degree of detail and nuance.
The Constitutional Court has, for example, not specified which procedure is to be followed when deciding on an ultra vires act or a derogation of national identity. Nor did it mention whether it would grant the European institutions any Fehlertoleranz, and whether this degree of tolerance would apply to both ultra vires acts and national identity derogations, or whether national identity would remain as inviolable as in the Bundesverfassungericht’s case law, allowing no leeway at all. Likewise, it did not delve into the delicate matter of what constitutes Belgian national identity. Analysis of these aspects demonstrates that the doctrine developed in judgment no. 62/2016 poses little threat to the full effect and uniform application of EU law.
Procedure
Although the Constitutional Court did not further elaborate on this point, it is very unlikely that it would decide upon an ultra vires act or an infringement of Belgian national identity without engaging in prior preliminary dialogue with the Luxembourg Court. The German Bundesverfassungsgericht, which had never referred a case to the European Court of Justice for a preliminary ruling, announced such a preliminary dialogue in Honeywell, and put it into practice in Gauweiler. Given that the Belgian Constitutional Court already possesses vast experience in referring cases to the Luxembourg Court for a preliminary rulingFootnote 71 and usually embraces its European mandate, it is safe to predict that it will follow the same procedural path. To be admissible, such a question must be framed in terms of the interpretation or the validity of EU law, and not as a question about Belgian law. As the Constitutional Court defines the concept of national identity by adopting the text of Article 4(2) TEU,Footnote 72 such a phrasing would not be too difficult: the Constitutional Court can ask for the review of an act of secondary EU law against Article 4(2) TEU, while explaining why the constitutional provision at hand amounts to an element of Belgian national identity.
Such a preliminary dialogue would allow the Constitutional Court to tackle the main criticism voiced against its current dialogue with the European Court of Justice. Some authors have questioned the Constitutional Court’s focus on technicality and docility in arguing and formulating its preliminary references, preferring that the Court would emphasise the particular constitutional principles at stake, especially when they are jeopardised by norms of EU law.Footnote 73 National identity discussions offer an interesting forum in which to do so.
It is quite likely that, after receiving the Luxembourg Court’s answer, the Constitutional Court would perform its subsequent ultra vires or national identity review europarechtsfreundlich and zurückhaltend.Footnote 74 A violation of the core of the Belgian Constitution or the attributed competences will probably only be sanctioned if it is a sufficiently qualified one,Footnote 75 and the Constitutional Court would probably grant the Luxembourg Court a significant degree of Fehlertoleranz.
Operationalising ultra vires or identity review and accepting the consequences
A more difficult question that needs answering is how the Constitutional Court can operationalise this review, and what the legal consequences would be of the Constitutional Court’s declaration that a piece of EU law was adopted ultra vires or had encroached on Belgian national identity. Regarding operationalisation, it should be noted that the Belgian Constitutional Court can only annul, or declare unconstitutional, Belgian legislation, whereas it cannot directly sanction European treaties or European legislation. A distinction needs to be made between the review of primary EU law and secondary EU law. Ultra vires review may only concern secondary EU law, whereas national identity review, and basic constitutional protection review, can concern both primary and secondary EU law.
It is rather difficult to submit primary EU law to the scrutiny of the Constitutional Court. Insofar as Belgian legislation approves a ‘constituent EU treaty’, it cannot be submitted to the Constitutional Court’s scrutiny in a preliminary reference.Footnote 76 Actions for the annulment of legislation approving treaties are possible, albeit within the shortened period of 60 days following the Act’s publication in the Moniteur belge.Footnote 77 The annotated judgment, however, shows that it is not self-evident that a citizen be granted standing in such a case.
Legislation transposing or implementing secondary EU law, by contrast, falls under the Court’s normal review competences. An action for annulment is admissible ratione temporis if it is launched within 6 months following the Act’s publication in the Moniteur belge. Preliminary references are not bound by any time limit, irrespective of the piece of legislation’s link to secondary EU law. The annulment of such legislation implies that the act of secondary EU law at hand is ex tunc deemed not to have been transposed or implemented.Footnote 78 A declaration of unconstitutionality in a preliminary ruling implies that, although this piece of legislation is technically still present in the Belgian legal order, judges are no longer allowed to apply it.Footnote 79
Although these consequences are not entirely equivalent to declaring a piece of EU law inapplicable, the effect is rather similar. Given the Constitutional Court’s cautionary approach concerning the full effect of EU law, it is very unlikely that it would take such a decision frivolously. From a practical point of view, the judgment under scrutiny should therefore mainly be read as a statement that EU law and European Court of Justice jurisprudence should not be adopted ultra vires and should not encroach upon the core of the Belgian Constitution, rather than as a real threat that the Constitutional Court will, any time soon, be jeopardising the full effect or uniformity of EU law.
Moreover, a judgment declaring the legislative approval, transposition or implementation of EU law unconstitutional would expose the Belgian State to political or legal sanctions at the international level. If the legislative approval of an EU treaty were annulled, the federal or federated Executives would still be bound by it at the international level, but the treaty would no longer have any effect within the domestic legal order. Such a situation would put the international relations of the Belgian State under pressure. If the legislative provisions transposing or implementing an act of secondary EU law were annulled, the EU’s sanction mechanism, including infringement procedures and damages claims, could come into play. If the Constitutional Court adopted such a judgment, it would thus need to accept that sanctions could follow. Rejecting federal policy by refusing to apply it while accepting beforehand all sanctions attached to non-compliance has recently been described as ‘uncooperative federalism’.Footnote 80 If, however, a constitutional court decides to act in an uncooperative manner in the EU context, these sanctions are not imposed upon the uncooperative body itself, but upon the member state whose elected policy makers do not necessarily agree with the Constitutional Court’s decision. This raises the question of whether a strategy of non-cooperation should be decided upon by an independent and therefore unaccountable court, or, rather, by a more democratically legitimised and accountable legislative or executive body.
Many European constitutional courts have answered this question by accepting their competence to perform an ultra vires review and an identity review.Footnote 81 This means that they see it as an important part of their constitutional mandate to protect the core of the national constitution against excessive intrusions by the EU. Acknowledgment of this mandate implies a willingness to sanction violations, even if doing so might backfire on the member state’s own interests.
In turn, this raises the more fundamental and highly disputedFootnote 82 question as to whether the principle of national identity is justiciable before national courts at all. It is clear that Article 4(2) TEU is indeed ‘law’ and that the ECJ must take it into account when interpreting EU law and when reviewing secondary EU law.Footnote 83 As an autonomous notion of EU law, the provision’s interpretation and application are, in principle, reserved for the Luxembourg Court, and not the national constitutional courts. However, it should also be noted that this provision only anchors the European perspective on national identity, whereas the national perspectives on national identity (often flagged as ‘controlimiti’ or ‘constitutional identity’) are to be found in the national constitutions and legislation. Interpreting and applying these national norms falls outside the Luxembourg Court’s mandate.Footnote 84 Both accounts of national identity being equally relevant in an environment of constitutional pluralism, application of the identity clause thus requires genuine dialogue between the national constitutional court and the Luxembourg Court. In its judgment no. 62/2016, the Belgian Constitutional Court certainly did not reject this dialogue, which already forms part of its routine when applying EU law.Footnote 85 Moreover, in defining the concept of national identity by referring to the text of Article 4(2) TEU, the Constitutional Court indicates that it understands its role in the dialogue.Footnote 86
Belgian national identity
A most delicate question is what exactly constitutes Belgian national identity. Identifying this notion is a daunting task for any State, given the concept’s vagueness,Footnote 87 but it is even more so for a bipolar State, in which the linguistic border also tends to divide public opinion. It therefore comes as no surprise that the Constitutional Court gives no indication as to the scope of Belgian national identity.
Grewe has suggested a method for identifying what constitutes a State’s national identity, based on the text of Article 4(2) TEU. According to her, constitutional provisions are likely to belong to the national identity if they are more difficult or impossible to amend or if they are closely linked to the Constitution’s preamble or introductory title.Footnote 88 These indicators, however, do not apply to Belgium, as the Belgian Constitution contains only one procedure for amendment, and does not contain an Ewigkeitsklausel, a preamble or an introductory title.
A State’s national identity can, moreover, not be identified by applying a provision of EU law, but solely by examining the particular constitutional history and culture of that country. It should also relate to aspects which are unique to that particular country.Footnote 89 Belgium’s constitutional history is characterised, on the one hand, by the specific reasons for its becoming an independent country, and, on the other, by the reasons which made it survive its linguistic and ideological crises which persist until this day. These episodes have clearly left their mark on the Constitution’s text and daily application.
First, Belgium’s 1831 Constitution clearly reflects the reasons which led to the Belgian revolution of 1830, i.e. the rejection of the Dutch King Willem I’s policies (1815-1830). He had usurped the residual powers, enabling him to prescribe the entire socio-economic policy by Royal Decree. Additionally, he impeded judicial review of these administrative acts. Furthermore, he adopted a strict linguistic and religious policy, which led him to impose severe restrictions on the freedom of religion, the freedom of language and the freedom of education. In order to dampen criticism, he also restricted the freedom of the press and the freedom of assembly.Footnote 90
The Belgian Constitution therefore strongly emphasises the ‘legality principle’. Several domains, such as tax, the military, the police, and the limitation of most human rights, have been explicitly reserved for the Legislature. The King has no political responsibility and all his acts require ministerial responsibility (Article 88 of the Constitution). He has no powers other than those explicitly attributed to him, while the residual powers reside with the legislator (Article 105 of the Constitution). The Judiciary is independent from the Executive, and all judges are granted explicit competence to review the legality of any Executive action (Article 159 of the Constitution). Finally, the human rights which were limited under the Dutch regime have been very rigidly guaranteed in Title II of the Constitution, some of which still benefit from a more extensive human rights protection than their counterparts in the Convention and the Charter: the freedom of the press, the freedom of assembly, the freedom of religion, the freedom of education and the freedom of language.Footnote 91 Most constitutional rights may only be limited by regulating their exercise or by sanctioning offences committed during their exercise, whereas preventive measures are generally forbidden.
The second set of constitutional features which arguably belong to Belgian national identity reveal another particularity of the Belgian polity, i.e. its talent for reaching compromise.Footnote 92 Belgian history shows several examples of bipolar opposition, such as the ideological opposition between Catholics, on the one hand, and Socialists and Liberals, on the other hand, culminating in the ‘School Issues’, which were resolved by the School Pact in 1958,Footnote 93 and the linguistic opposition between the Flemish and the Walloons, starting as early as the 1840s, culminating in violent student protests in 1968 and eventually leading to Belgium’s transformation into a federal state sui generis. These fundamental compromises were reached after difficult negotiation and all their aspects are of equal importance to the pacifying compromise. Therefore, these compromises must be considered part of Belgium’s ‘fundamental structures, political and constitutional’. If any element of this delicate compromise were to perish under EU scrutiny, the whole equilibrium would be lost.
Therefore, the basic choices made during Belgium’s federalisation process can be considered to make up part of its national identity.Footnote 94 In this respect, the fundamental choices for territorial federalism, the rule of linguistic parity, and the specific linguistic regulationsFootnote 95 are crucial. The specific choices typical of Belgian federalism regarding the operation of participative and cooperative federalismFootnote 96 and regarding minority protection at the group level rather than at the individual level must also be mentioned.
The specific characteristics of human rights protection under the Belgian Constitution offer interesting prospects viz the Luxembourg Court’s Melloni jurisprudence.Footnote 97 Given that both the substance of several constitutional rights and the procedure for their limitation belong to Belgian national identity, the principle of the most extensive human rights protection is – although not explicitly mentioned in the Constitution – of critical importance. Arguably, as Belgian constitutional rights belong to Belgian national identity insofar as they offer more extensive human rights protection than the Convention and the Charter, the EU institutions, including the Luxembourg Court, may not in any way jeopardise constitutional review concerning these rights. Hence, identity review might prompt the Constitutional Court to set aside the Melloni judgment and to apply the full ambit of constitutional rights, even within the scope of EU law. In this context, it is useful to remember that judgment no. 62/2016 not only mentioned national identity as such, but also ‘the basic values of the protection offered by the Constitution to all legal subjects’, which can be thought of as a separate standard or as part of the national identity.Footnote 98
Discriminatory derogation to the national identity
Finally, attention should be paid to the way the Constitutional Court seemingly qualifies national identity review by only disallowing ‘a discriminating derogation from the national identity’. This referral to the principle of equality, which does not occur in the Bundesverfassungsgericht’s constitutional identity case law, seems odd. If derogation from national identity is sufficiently serious to raise a shield against the applicability of EU law, it should not matter whether the derogation is discriminating or not. Moreover, elements of national identity are always rather specific to one Member State, and a derogation to national identity will therefore always single out one or two Member States.
The mention of the principle of equality is probably reminiscent of the technique used by the Constitutional Court to grant itself the competence to review legislation against EU law.Footnote 99 This technique, however, is irrelevant when EU law – or its legal implementation or transposition – is itself under scrutiny.
Conclusion: uniform application of EU law in jeopardy?
In her PhD thesis, Cloots stated that ‘national constitutional scrutiny has hardly ever done any direct harm to a measure of secondary EU law’.Footnote 100 The same applies to the Belgian Constitutional Court’s judgment analysed in the present case note. Although the Constitutional Court’s judgment no. 62/2016 can be read as a rejection of an automatic application of the principles of full effect and uniform application of EU law, it remains doubtful whether the Court will put the review it announced in practice any time soon. Several reasons prompt this conclusion.
First, the Belgian Constitutional Court remains one of the most Europhile jurisdictions, applying EU law and referring cases to the ECJ on a regular basis. Judgment no. 62/2016 cannot be read as a repudiation of that approach.
Second, although quite succinct, the judgment does adhere to the Bundesverfassungsgericht’s long-standing and very elaborate case law concerning the interface between constitutional law and EU law. German case law has never caused great damage to the validity of norms of EU law within the German legal order. It is very likely that, if the Belgian Constitutional Court applied the new test, it would also adopt the built-in mechanisms of self-restraint, both procedurally and substantively: it would probably engage in preliminary dialogue before applying an identity or ultra vires review, or before enforcing the basic values of the protection offered by the Constitution, and would probably apply any further review zurückhaltend, thus granting the Luxembourg Court a significant degree of Fehlertoleranz.
Finally, the Constitutional Court’s vagueness on the procedural and substantive aspects of its new test, and on the content of Belgian national identity, is indicative of its ambition. Judgment no. 62/2016 was probably not meant to be operationalised any time soon, but should rather be understood to be an expression of the Court’s discontent with the current evolution of the Luxembourg Court’s case law concerning the full effect and uniform application of EU law. The Melloni judgment’s limiting the ambit of constitutional review might have been the straw that broke the camel’s back.
Nevertheless, despite the limited practical relevance, the Constitutional Court did choose to make such a statement. In a setting of multilevel constitutionalism in which both direct and indirect judicial dialogues are crucial, the Luxembourg Court should pick up on such signals and evaluate them when deciding upon the possible further evolution of its full effect case law.
