The Court of Justice in times of politicisation: ‘law as a mask and shield’ revisited

ABSTRACT

This contribution analyses if and under what conditions bottom-up pressures constrain the Court of Justice of the European Union (CJEU). Neofunctionalists famously explained the power of the Court by its use of ‘law as a mask and shield’. Due to its technical nature, the Court is able to mask the political substance of ‘integration through law’ and to shield it from political challenges. We revisit this argument in times of politicisation of the EU and develop a typology of four constellations depending on different kinds of bottom-up pressures. We argue that depoliticised integration through law still functions, to varying degrees, in face of either governmental contestation or public politicisation. By contrast, if member state governments and the public jointly oppose further integration through law, the Court is constrained as further expansive jurisprudence could even reinforce politicisation and risk political backlash. We illustrate our argument with case law on EU citizenship.

KEYWORDS:

• Court of Justice
• EU citizenship
• integration through law
• Neofunctionalism
• politicisation

Introduction

Among all major EU institutions, the Court of Justice of the EU (CJEU) is the least likely to be affected by the bottom-up politicisation of EU decision-making. The Court has been famously described as ‘tucked away in the fairyland Duchy of Luxembourg’ (Stein, 1981, p. 1) to stress the remoteness of European judges from EU political battles, let alone broader public attention. As with any court, the CJEU’s authority rests on its perception as a non-political actor, settling conflicts by technical-legal means, and hence, CJEU judges have no interest in becoming agents of politicisation. Rather, the Court’s ability to depoliticise conflicts is a crucial precondition for its role as an engine of integration. Due to its technical-legal nature, neofunctionalists have argued, law serves as a ‘mask and shield’, which obscures the political substance of CJEU jurisprudence and protects it from political challenges.

And yet, the CJEU is not immune against bottom-up pressures. Escaping the ‘dictate’ from Luxembourg was a major issue during Brexit negotiations. More generally, research on courts and politics has shown that judges do not rule in a political vacuum, but are responsive to changes in the broader ‘political mood’(McGuire & Stimson, 2004). Against the background of increasing EU politicisation, we therefore revisit Burley and Mattli’s classical argument about ‘law as a mask and shield’.

We develop a typology of four constellations depending on whether CJEU jurisprudence is subject to bottom-up pressures stemming from member state governments and/or politicisation. Based on our distinction of bottom-up pressures, we qualify the ‘mask and shield’ argument in two respects. First, even if ‘shielded’ against outright political attacks, CJEU jurisprudence remains vulnerable to governmental contestation by legal means. Integration through law scholars underestimated member states’ ability to contain the impact of expansive CJEU jurisprudence through creative compliance at the domestic level and legislative corrections at the EU level. Second, we argue that the Court faces a ‘constraining consensus’ if further expansive jurisprudence is not only contested by member state governments but faces also public politicisation. Under these circumstances, judges have to choose between self-restraint and the risk of political backlash (Madsen, Cebulak, & Wiebusch, 2018).

The following theoretical section briefly summarises the original neofunctionalist account, before we develop our own argument and typology of different constellations of CJEU contestation. Subsequently, we illustrate these four ideal-typical constellations with empirical evidence from the area of EU ideal-typical complemented with examples from other policy areas. The concluding section discusses the prospects of further Court-driven integration through law in times of politicisation.

Law as a mask and shield revisited

The original argument

In their seminal article ‘Europe before the Court’ from 1993, Anne-Marie Burley and Walter Mattli applied neofunctionalist theory to explain the CJEU’s strength as an engine of integration (Burley & Mattli, 1993). The Court, they argued, often ruled in favor of further integration to an extent that goes significantly beyond what could be reached by the political agreement of governments alone. Crucially, this integration through law depends on the distinctiveness of the legal domain:

Law functions as a mask for politics, precisely the role neofunctionalists originally forecast for economics. The need for a ‘functional’ domain to circumvent the direct clash of political interests is the central insight of neofunctionalist theory. This domain could never be completely separated from the political sphere but would at least provide a sufficient buffer to achieve results that could not be directly obtained in the political realm. (Burley & Mattli, 1993, p. 44)

Burley and Mattli linked their argument already explicitly to (de)politicisation: ‘The strength of the functional domain as an incubator of integration depends on the relative resistance of that domain to politicisation’ (Burley & Mattli, 1993, p. 44). Court judgments appear as ‘technical’, they deny the ‘existence of policy discretion’ and provide ‘nonpolitical justification(s)’ for ‘purportedly neutral’ decisions, which derive from their ‘distinctive methodology and logic’, in short: they follow the ‘inexorable logic of law’ (Burley & Mattli, 1993, p. 69–73). Due to these characteristics, case law ‘functions both as a mask and shield’ (Burley & Mattli, 1993, p. 74). The political implications of judgments are hidden behind their technical-legal language. And as long as no-one wants to call the Court’s general authority into question, individual rulings are protected from outright political contestation, i.e., politically unwelcome case law can only be challenged on its own legal terms.

While Burley’s and Mattli’s argument deals with the depoliticising effect of law and courts in general, it applies, particularly, to the EU context. Appointments to the CJEU are much less prone to politicisation than, for example, to the US Supreme Court (Ferejohn, 2002, p. 65), and the selection process of CJEU judges has been largely professionalised on the basis of the Lisbon Treaty (Dumbrovsky, Petkova, & Van der Sluis, 2014). Moreover, apart from efficiency considerations, internal decision-making at the CJEU remains largely shielded from public scrutiny. While greater openness and transparent reasoning might arguably improve the legal quality of the Court’s jurisprudence (Weiler, 2013), European judges have always cared greatly about maintaining the non-political appearance of their case law, e.g., through the short and apodictic style of many judgments, the confidentiality of their internal decision-making and the absence of dissenting opinions.

Two qualifications

25 years after its original publication, the ‘law as a mask and shield’ argument is still one of the most prominent accounts of CJEU-driven integration through law; arguably holding a similar status as standard reference among political scientists as Joseph Weiler’s ‘Transformation of Europe’ (1991) in the legal field. Yet, the neofunctionalist argument is not uncontested. Empirically, it has been noted that Burley and Mattli offered little evidence for their argument beyond the evolution of CJEU legal doctrine itself (Stone Sweet, 2010, p. 16). Subsequent research delivered strong support for some of their claims, e.g., regarding the enabling role of subnational actors such as domestic courts, lawyers and private litigants (Alter, 1998a; Cichowski, 2004; Kelemen, 2011; but see also Conant, Hofmann, Soennecken, & Vanhala, 2018 for many open questions regarding EU legal mobilisation). Theoretically, historical institutionalists explained the acceptance of expansive CJEU jurisprudence alternatively by emphasizing the short time horizons of politicians rather than the inherent qualities of law (Alter, 1998b; Pierson, 1996). Moreover, it has been criticised that neofunctionalists were largely ignorant of the limitations of integration through law (Alter, 1998b, p. 121). And indeed, Burley’s and Mattli’s original account shows only little concern for potential qualifications of the ‘mask and shield’ argument: in order to be accepted as law, CJEU jurisprudence just needs to exhibit ‘a minimal degree of fidelity to both substantive law and the methodological constraints imposed by legal reasoning’ (Burley & Mattli, 1993, p. 44) and in order not to lose its overall legitimacy, the Court must not ‘outrun its constituency’ (Mattli & Slaughter, 1998, p. 181).

In the following, we do not seek to present an alternative account, but we introduce two qualifications concerning both the shield and the mask aspects of CJEU jurisprudence, which are inspired by recent research on (i) judicial impact and on (ii) politicisation.

First, the shield of EU law is not impenetrable. When confronted with unwelcome case law, EU member state governments typically do not challenge the Court’s overall authority politically, but they may seek to limit the Court’s impact by legal means. Burley and Mattli captured judicial impact only partially. Although they defined their dependent variable in terms of a ‘gradual penetration of EC law into the domestic law of its member states’ (Burley & Mattli, 1993, p. 43), their empirical analysis focused on the development and acceptance of CJEU legal doctrine rather than on broader policy impact. Probably the most radical way to limit the Court without questioning its existing authority is the exclusion from new areas of integration (Burley & Mattli, 1993, p. 74). Apart from exclusion, a growing body of research on CJEU judicial impact has analyzed member states’ more subtle ways to control and limit the effects of expansive jurisprudence. At the European level, member states may adopt secondary legislation to modify the implications of existing and to preempt future case law. Whereas outright legislative override of CJEU jurisprudence is rare, member states have been found to draft secondary legislation carefully in order to circumscribe the Court’s policy impact (Martinsen, 2015). At the domestic level, the impact of unwelcome Court jurisprudence may be effectively limited in the process of implementation. As with legislative override at the EU level, member states rarely opt for open non-compliance at the domestic level. Rather, they resort to creative forms of compliance, e.g., by limiting the Court’s impact to the individual conflict at issue (Conant, 2002) or by carefully designing domestic reforms ‘with Luxemburg in mind’, which are likely to survive future lawsuits while minimizing substantive policy change (Blauberger, 2012).

Secondly, the political substance of the Court’s case law has become harder to mask. Burley and Mattli wrote their original piece still under the impression of the ‘permissive consensus’ (Hooghe & Marks, 2009). Since then, politicisation of European governance has increased in terms of visibility, polarisation and engagement beyond elite actors (Bresanelli, Koop, & Reh, 2020; de Wilde, Leupold, & Schmidtke, 2016). And, closely linked to this overall trend, CJEU jurisdiction has come to involve ever more politically sensitive issues such as citizenship or welfare rights, which forces the Court to engage in delicate value judgments and interest evaluations (Kelemen, 2016, p. 126f.). As a consequence, even when adhering perfectly to legal methodology and existing doctrine, CJEU judges may be confronted with politicisation. Particularly, the visibility or salience dimension of politicisation is important for our discussion of bottom-up pressures on the Court. Arguably, actor expansion might rather reinforce EU legal mobilisation (cf. Kelemen, 2011) and polarisation might increase CJEU independence as it reduces the risk of politically agreed counter-action (Larsson & Naurin, 2019, p. 12). By contrast, if politicisation involves salient public opposition against further integration in an area of CJEU jurisdiction, it can have a constraining effect on the Court. Studies on domestic high courts have shown that public opinion places a constraint on judicial decision-making and that judges respond to changes in the broader political context (McGuire & Stimson, 2004). Recent research on international courts also emphasises the importance of contextual factors for judicial authority (Alter, Helfer, & Madsen, 2018), including diffuse public support or opposition. Even if the vast majority of CJEU judgments do not reach broad visibility, newly collected data on newspaper coverage of CJEU jurisprudence shows that a considerable share of case law is in fact reported by main European newspapers (Dederke, 2018, p. 18). Moreover, there is empirical evidence that CJEU actually do ‘read the morning papers’ (Blauberger et al., 2018).

Four constellations of CJEU contestation

Following from the preceding discussion, we distinguish two kinds of bottom-up pressures for the Court. On the one hand, a large part of the political science literature on the CJEU takes a state-centric approach and addresses the Court’s relationship to member state governments. Essentially, the question is how far CJEU judges can go in promoting integration through law and making it effective even against opposition from member state governments. While we cannot define a precise cut-off point of how many member governments constitute a bottom-up pressure on the Court, governmental contestation is typically regarded to be greater, the more member governments oppose an expansive EU legal interpretation in their written observations to the Court (Carrubba & Gabel, 2015, p. 84; Larsson & Naurin, 2016, p. 396). On the other hand, we add public politicisation as a separate kind of bottom-up pressure potentially affecting the Court’s independence and impact.

Combining these two kinds of bottom-up pressures, four constellations of Court contestation are possible (see Table 1) and each constellation has different implications for the Court’s strategic responses and their effectiveness in promoting integration through law. In constellation I, the Court is largely unconstrained by governmental contestation and public politicisation: integration through law is effective even without particular depoliticising efforts. In constellations II and III, the Court is faced with bottom-up pressures from either public politicisation or governmental contestation: both, ‘law as a mask’ and ‘law as a shield’ are responses of ‘assertive depoliticisation’ (Bressanelli et al., 2019), i.e., they enable the Court to rule expansively despite bottom-up pressure. Finally, in constellation IV, public politicisation and governmental contestation constitute a ‘constraining consensus’ for the Court: a strategy of judicial self-restraint is necessary to protect the Court’s recognition as a non-political institution and to avoid political backlash more generally.

Table 1. Four constellations of CJEU contestation.

	Public politicisation
		No	Yes
	No	(I)	(II)
Governmental contestation		Permissive consensus	Law as a mask
	Yes	(III)	(IV)
		Law as a shield	Constraining consensus

(I) Unsurprisingly, the Court is least constrained in its role as an engine of integration in the absence of both, governmental contestation and public politicisation. Arguably, this constellation has predominated for the long period of the ‘permissive consensus’ and in particular in the core area of EU single market integration. The public exhibited a permissive attitude to elite-driven integration in general and largely ignored the ‘quiet revolution’ (Weiler, 1994) with the Court as the ‘unsung hero’ of integration (Burley & Mattli, 1993, p. 43) in particular. Member state governments disliked some rulings, but generally welcomed the Court’s role in overcoming political obstacles to European market integration. Even if individual judgments were met with opposition from one or two governments, this hardly constrained the Court. As disagreement over the direction and contents of law are defining features of law-making (Madsen et al., 2018, p. 202), contestation by individual member states hardly constrains the court or challenges its authority – but merely constitutes business as usual (Hofmann, 2018; Larsson & Naurin, 2016, p. 392).

(II) Burley’s and Mattli’s original argument about law as a mask mainly referred to the Court itself ‘masking’ the political implications of its jurisprudence. We amend this argument in light of our two qualifications above. Given that member state governments have become more alert to Court-driven integration over time, they can unmask unwelcome integration through law more readily. However, they may prefer and protect the mask of law in light of public politicisation, i.e., whenever they share the Court’s interest in promoting integration, but shy away from taking the political responsibility for unpopular decisions. As a consequence, the function of law as a mask is most powerful when the Court and member state governments collude. Under these circumstances, governments may prefer to shift decision-making to the seemingly neutral area of the judiciary, while tacitly agreeing with the political substance of Court-driven integration through law. By arguing ‘We are sorry, but it’s the law’, member state governments can at least partly shift the blame to the Court. At the same time, the Court’s authority remains unquestioned and member state governments do not undermine the impact of its jurisprudence.

(III) The Court can use law as a shield most forcefully, when governments are opposed to a specific ruling, but their opposition is not paralleled by strong public politicisation of that issue. In this constellation, the Court has a particular interest in delivering high quality judgments, which are difficult to challenge on legal grounds. References to established precedent serve to emphasise overall legal coherence and certainty (Schmidt, 2012, p. 10) and to justify further expansionary jurisprudence: ‘where the “mask of law” has already fallen … the Court argues more carefully, by means of reference to precedent’ (Larsson, Naurin, Derlén, & Lindholm, 2016, p. 881). Under these circumstances, any attempt to deliberately politicise and to openly resist the Court’s jurisprudence is likely to backfire for member state governments. Throughout the union, citizens tend to have greater trust in the CJEU than in their own government (or, actually, any other EU institution, see Kelemen, 2012, p. 48f.). Moreover, even if governments are very unhappy with an individual judgment, questioning the general authority of an otherwise functional court would be a high price to pay. Consequently, they seek ways to limit the impact of CJEU jurisprudence by legal means, e.g., through legislative modification at the EU level or creative compliance domestically (Hofmann, 2018). Such measures may contain the impact of individual rulings considerably, but by countering the case law on its own legal terms, they do not call into question the Court’s overall authority.

(IV) Finally, CJEU judges may become confronted with the fourth constellation: a ‘constraining consensus’ where governmental contestation occurs together with public politicisation. While ‘law as a mask’ crucially depends on the collusion of governments and while law ‘shields’ the Court from political attacks without public resonance, both protective mechanisms fail in face of joint opposition from member state governments and the general public. This constellation occurs rarely since CJEU judgements usually do not reach broad visibility while at the same time being collectively opposed by several governments. Such bottom-up pressure is most likely to arise when the Court develops innovative rulings against the expressed intent of the EU legislature which affect matters of high political significance (Hofmann, 2018, p. 260; Madsen et al., 2018, p. 205). If, however, the Court continues to drive integration against such a ‘constraining consensus’, it risks being perceived as a political actor who can also be challenged in political terms. Contestation, then, will not just aim at reducing the impact of individual rulings, but target the Court’s general authority to rule expansively. Under these circumstances, the technical complexity of case law may even have the inverse effect of what neofunctionalists described, i.e., populists may (ab)use the opaque and complex character of individual cases and paint ‘law as a caricature’ to make vastly exaggerated anti-EU claims. As a consequence, we would expect CJEU judges to back down in the face of a ‘constraining consensus’ and to limit or even reverse expansive jurisprudence. Even if this scenario occurs only rarely, its relevance may be broader, namely if CJEU judges exert self-restraint in anticipation of possible backlash.

In sum, our typology refines the original mask and shield argument by specifying the constellations under which each of these functions of law plays a particular role and by adding public politicisation as an important contextual factor for European judges. It implies one basic hypothesis: effective integration through case law becomes less likely as we move from constellations (I) to (IV). Apart from the law itself, the Court is least constrained (I) under a permissive consensus and (II) when law serves as a mask. The Court may (III) continue to rule expansively even against member state opposition and use law as a shield against outright political attacks, but the impact of expansive jurisprudence is limited. Finally, (IV) under the rare constellation of a constraining consensus, the Court is limited not only with respect to its impact, but in its ability to rule expansively in the first place.

Empirical illustration

In the following, we illustrate these four ideal-typical constellations with empirical evidence from a joint project on the free movement and equal treatment of EU citizens, complemented with well-researched examples from other policy areas. We, thereby, focus on a rather extreme set of cases, which are not meant to ‘test’ causal hypotheses or ‘represent’ CJEU case law in general. Instead, these cases are particularly useful for illustrative purposes as they come close to the ideal types conceptualised above and they cover the full spectrum of our typology. First, while case law on the free movement of workers reaches back to the earliest phases of integration and of the permissive consensus, it is arguably the area in which the Court ruled most strongly against member state opposition (Larsson & Naurin, 2016, p. 396). Secondly, while many CJEU judgments do not reach any public attention (Dederke, 2018, p. 18), the free movement of EU citizens and especially their cross-border access to social welfare has become increasingly politicised in recent years. We have thus chosen a set of cases where legal integration has occurred for long and considerably, while at the same time politics is a play (Martinsen, 2015, p. 10).

(I) Permissive consensus

In terms of producing cases, the CJEU has been remarkably effective. Between 1961 and 2014, the Court decided in more than 7000 preliminary rulings and infringement proceedings (Martinsen, 2015, p. 11). Most cases have concerned agriculture and fisheries, second comes internal market interpretations concerning free movement of goods and thirdly judicial decisions within the social realm. Arguably, the largest pile of these cases were never politically controversial, but merely rather technical legal disputes, relevant for the parties involved but without broader implications. However, famous cases such as Van Gend en Loos and Costa v. E.N.E.L. from 1963 to 1964 certainly fall in this category of integration through law and were enabled by a permissive consensus. New historical research has shown that these landmark cases did indeed cause some political concern, but the majority of governments did not mobilise against the new legal order constructed (Rasmussen, 2017).

Expansive CJEU jurisprudence has largely shaped the free movement and equal treatment of workers for decades, enabled by the permissive consensus. Originally, the right to cross-border welfare in the European Union was restricted to workers by means of Regulation 1612/68 (now 492/2011) and Regulation 1408/71 (now 883/2004), entitling Community migrant workers to equal treatment with national citizens on social rights and tax advantages next to exporting social-security rights across member-states borders. This original regulatory framework of the 1960s has gradually extended its personal and material scope by means of CJEU case law and legislative revisions. The Court’s early jurisprudence laid down that the purpose of the Treaty was to establish as ‘complete a freedom of movement for workers as possible’ (Case 75/63 Hoekstra). It later extended the personal scope to also include the self-employed (Case 17/76 Brack) and clarified the rights of the family members (Case 7/75 Fracas; C-308/93 Cabanis-Issarte). In addition, a permissive consensus allowed the Court to develop an inclusive concept of who is a worker according to EU law. In the case of Kempf (C-139/85), the CJEU established that working 12 h per week would suffice, and in the case of Megner and Scheffel (C-444/93), it ruled that 10 h of work per week did not exclude a person from being regarded as a worker.

EU gender equality has been another often cited cluster of case law, used to illustrate the power of integration through law (Cichowski, 2004). The Defrenne cases from 1971 and 1976 were crucial in defining the meaning and effect of the Treaty’s declaration on equal pay and initiated an important process of judicial and political integration, which proved that litigation can be an important tool to push for change. In the Defrenne II case from 1976, a Belgian Court asked if the Treaty’s article 119 had direct effect, creating enforceable rights for individuals in national courts without EU secondary legislation or national legislation transposing it. The Court disregarded objections by the UK and Ireland and confirmed the direct effect of article 119. Other important cases followed as for example Danfoss (C-109/88) on equal pay and Dekker (C-177/88) concerning the rights of pregnant workers among many others. The Court’s jurisprudence was later codified into important pieces of secondary legislation (see for example the burden of proof directive 97/80 adopted in 1997). A large part of gender equality case law occurred without any significant governmental contestation despite their political and also financial implications. Although some member states disliked the rulings, there was no broader collective mobilisation or public politicisation for or against them.

(II) Law as a mask

While law may hide political conflict in various ways, e.g., when the broader and long-term consequences of individual rulings go unnoticed by member state governments, the aspect of law as a mask is essential when governments tacitly endorse integration through law, but do not want to overtly take political responsibility. Theoretically, such a scenario in which governments shift decision-making (and blame) for unpopular decisions is highly plausible, but empirically, these cases are inherently difficult to document. The Court’s role in these cases lies precisely in serving as a forum in which governments can hide politically conflictual issues behind a mask of technical, legal reasoning. And governments are assumed to remain silent or even to delude the public about their actual preferences.

Nevertheless, we can draw on some well-researched examples from the posting of workers to illustrate this constellation. The question of which labor standards apply to workers, who are posted from one EU member state to another, was at the heart of the ‘Laval quartet’ (Blauberger, 2012). In three of these four landmark judgments, the Court interpreted the Posting of Workers Directive extensively and ruled against national restrictions of the free movement of services (Martinsen, 2015). The cases were mainly perceived as pitting old against new EU member states, with the CJEU favoring further liberalisation and, thus, siding with new EU member states (Lindstrom, 2010). And yet, there is considerable evidence that the Court rulings were also received with the tacit approval of some conservative-liberal governments in the old member states. The Rüffert judgment (C-346/06) had the greatest impact in those German regions governed by conservative-liberal coalitions, where it was used to abolish the contested domestic rules without replacement (Sack, 2012). The Laval judgment (C-341/05) triggered various domestic reform processes, which went furthest in Sweden, where a conservative-liberal government was under strong influence from major business interests (Seikel, 2015). Arguably, in these cases, the mask of EU law worked better for the governments involved than for the CJEU itself. Since these cases were already politicised when they reached the Court, European judges were hardly able to mask the underlying political conflicts and had to take inevitably unpopular rulings. By contrast, the Court’s rulings enabled liberal-conservative governments to enact domestic reforms, which could be portrayed as EU legal necessities.

Another policy area in which EU law typically serves to mask political conflict is the control of state aid. Over time, the Commission has been able to incrementally tighten the control of state aid in the EU, largely supported by the CJEU, but also with the tacit agreement of member state governments (Blauberger, 2009). Governments share a collective interest in avoiding subsidy races and reaping the benefits of undistorted competition, but they are vulnerable to political pressures from potential beneficiaries among their constituents. For example, governments are regularly called upon to save domestic enterprises (and their jobs) from bankruptcy by granting state aid. Under these circumstances, governments often readily refer to the constraints imposed by EU rules to deny state aid, while shifting the ultimate responsibility and blame to the European level.

(III) Law as a shield

The case law on the free movement of workers post-Maastricht offers plenty of evidence on how the Court’s jurisprudence is shielded from outright political attacks despite challenges by member state governments. Since Maastricht, the Court has significantly extended the rights to move freely and to be treated equally from workers to EU citizens more generally (Wollenschläger, 2011). The actual impact of the Court’s citizenship jurisprudence, however, has been more limited. Already before the Court has shifted its own course (see next section), member state governments have effectively limited its impact through legislative modification at the EU level and contained compliance at the domestic level. Modification means compromises between legislative actors where specific articles or annexes insert more national control, more national discretion, exemptions or special rules (Martinsen, 2015, pp. 35–36). Such mitigating articles confines the implications of the case law nationally and in its practical application. For example, the expansive Grzelczyk ruling (C-184/99) from 2001 laid down that Union citizenship entitles students from other member states to the social assistance of a hosting member state if a certain link with the society of their host state has been established. Soon after, however, the member governments constructed a delicate compromise with the European parliament in the Citizenship directive 2004/38 (Wasserfallen, 2010), where article 24.2 aimed to modify the implications of the Court ruling by stating that member states are not obliged to confer entitlement to study grants or student loans before permanent residence has been acquired. Modification thus expresses governmental contestation to Court rulings, but is still a milder form of political pushback than upright rebellion against the authority of the Court. Moreover, there is increasing evidence that the impact of the Court’s citizenship jurisprudence was limited from the beginning due to restrictive implementation at the domestic level (Heindlmaier & Blauberger, 2017; Martinsen, Rotger, & Thierry, 2019).

Moving beyond the area of EU citizenship, we also find examples of more severe measures to limit the Court’s impact by legal means. Most often, the threshold to agree on legislative override is too high for member state governments, which is why they resort to modification as a second best option. And yet, there are some examples of legislative override. Again gender equality is illustrative, where the political receptiveness of legal innovations ended abruptly with the adoption of the Maastricht treaty in 1993. The Court’s Barber case (C-262/88) from 1990 had laid down that the principle of equal treatment also applied to age of retirement. This time, member states and employers argued strongly against the Barber case’s conclusions and the member states unanimously adopted the Barber protocol in the Maastricht Treaty to limit the implication of the judgement. Governments may also try to weaken the relative importance of the Court for the integration process more generally. For example, they restricted the power of the Court when they excluded its jurisdiction from two of the three Maastricht Treaty pillars (Burley & Mattli, 1993).

(IV) Constraining consensus

Finally, the Court faces a ‘constraining consensus’ under very particular circumstances – namely, if the Court is continuously supplied with new cases in an area, in which further integration is opposed simultaneously by more member governments and public opinion. In this situation, the Court has to make decisions without, however, law serving as a mask or shield insulating the Court from governmental and public opposition. By extending integration through law regardless of political and public opposition, the Court would even exacerbate politicisation. As a consequence, the Court is likely to be responsive and to exert self-restaint in face of strongly voiced political and public objections.

In the area of free movement and cross border welfare, legal integration for many years advanced not only the social rights of workers but also of non-economically active EU citizens. Based on European citizenship enshrined in the Treaty, the Court was heralded for adding real rights and substance to an otherwise empty political concept. In core cases, such as Martínez Sala (C-85/96), Grzelczyk (C-184/99), Baumbast (C-413/99) and Trojani (C-456/02), the Court extended the principle of equal treatment to lawfully residing Union citizens, not on the basis of their economic activities but ‘purely’ as citizens of the Union. It thus declared that Union citizenship as ‘destined to become the fundamental status of nationals of the Member States’ (Grzelczyk, para. 31). However, during the last five years political and public critique against free movement and cross border welfare rights has sounded louder. Member state governments, not the least from the UK, had already called for greater national discretion regarding the regulation of EU migrants’ welfare access for several years. In addition, between 2013 and 2015, the public debate around the theme of ‘welfare migration’ increasingly politicised EU citizens entitlements to welfare in a hosting member state in several member states (Blauberger et al., 2018). ‘Welfare tourism’ came to feature as a negatively framed concept in the public debate. At the same time, the recent citizenship jurisprudence shows that the CJEU is not immune to the public temper. The previously expansive Union citizenship jurisprudence has turned into a remarkably restrictive course of legal integration with the cases of Dano (C-333/13) in November 2014, Alimanovic (C-67/14) in September 2015, Garcia-Nieto Garcia-Nieto (C-299/14) in April 2016 and in June 2016 with the Commission v. UK (C-308/14) infringement case. The Union citizenship of the treaty is no longer the reference point in the Court cases; instead the later restrictive interpretations are based on the conditions of secondary legislation as stated in the Citizenship directive. In its restrictive case law on free movement and cross border welfare, the Court not only proves attentive to the EU legislature, but sensitive to the broader political context and has turned away from expanding right of non-economically active EU citizens (Blauberger et al., 2018).

The ‘Laval quartet’ was already discussed above as it provided interested governments with a short-term opportunity to mask their own preference and to shift the blame for unpopular decisions to the CJEU. In the longer run, however, the Court could not sustain its expansive jurisprudence as it became confronted with a constraining consensus. The Court’s judgments triggered unprecedented public criticism (Scharpf, 2008) as well as resistance by a significant number of member states. In 2012, the Commission’s attempt to codify the Court’s case law in the so-called ‘Monti II Regulation’ led to the first activation ever of the Early Warning Mechanism by twelve national parliaments. When asked to further clarify its interpretation of the Posted Workers Directive in subsequent cases, the Court ‘readjusted its position to the benefit of national social regulatory autonomy’ (Garben, 2017, p. 38). In the case Elektrobudowa (C-396/13), the Court sided with the claims of a Finnish trade union and adopted an unexpectedly generous interpretation of what could be subsumed under the crucial term of ‘minimum rates of pay’. In response to the Rüffert judgment (C-346/06), most German regions revised their public procurement legislation in order to protect as much as possible of its original substance in an EU-compatible way (Blauberger, 2012, p. 118). When this legislation was challenged again in the Regiopost case (C-115-/14), the Court disagreed with the Commission and shifted its previous position: ‘To avoid an explicit and principled reversal of its case law, the Court confines the ruling to the specific facts (…) But even if the extent to which Laval, Rüffert, Luxembourg and Bundesdruckerei have been overturned is open to discussion, the permissive signal to Member States and their courts is a strong one’ (Garben, 2017, p. 39).

Finally, another politically sensitive area, in which the Court appears to exert self-restraint in anticipation of (rather than in response to) a constraining consensus, is the area of fundamental rights. Member state governments have been cautious not to facilitate unwarranted CJEU intervention into domestic affairs and limited the scope of application of the Charter of Fundamental Rights of the EU ‘to the Member States only when they are implementing Union law’ (Article 51 of the Charter). Obviously, any attempt to extend the Charter’s scope of application through case law hinges on the interpretation of when member states ‘are implementing Union law’ and the Court has raised some expectations that might adopt an expansive interpretation (e.g., in the Åkerberg Fransson case, C-617/10). Yet, in contrast to these expectations and despite numerous ‘invitations’ to do so, the Court has so far refrained from an extensive interpretation of Article 51 of the Charter. Apart from the (lack of) legal basis, the current CJEU judge von Danwiz justifies the Court’s restraint with a more or less explicit reference to the potential politicisation which could be caused by greater judicial activism (von Danwitz, 2014).

Conclusion

25 years have passed since Burley and Mattli first presented legal integration to function as a mask and a shield (Burley & Mattli, 1993). Since then, political awareness of the outcomes of integration through law has grown significantly and EU legislative actions as well as national counter-actions limit the political power of the Court. Member states have become more skillful in penetrating the shield of law and as EU integration has moved into core areas of national sovereignty, the political implications of jurisprudence have become harder to mask. Public opinion increasingly forms for or against EU policies in general and partly even for or against specific judicial interpretations. As a consequence, growing politicisation has become an additional scope condition for the CJEU to acknowledge and, under peculiar circumstance, a ‘constraining consensus’ may restrict the possible outcomes of integration through law, when public politicisation is paralleled by member state opposition to expansive CJEU jurisprudence.

The judges in Luxembourg are not immune to this ‘constraining consensus’. The CJEU appears to be sensitive to changes in the broader political context. This is not to say that the European Court has become or is likely to become politicised itself in the sense of taking a more ideological political side, but rather that public politicisation constrains the CJEU’s ability to promote integration through law. Ultimately, the Court’s legitimacy depends on its jurisprudence being accepted and complied with and there is only that much opposition a judicial authority can take. Moreover, once an issue has already become broadly politicised, we should not overestimate the Court’s ability to simply depoliticise it again by the mere force of law. Hence, despite the function of law as a mask and shield, growing politicisation as a scope condition is largely exogenously given for the Court.

Acknowledgements

We would like to thank Christine Reh, R. Daniel Kelemen and three anonymous reviewers for their very helpful comments.

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes on contributors

Michael Blauberger is Professor of European Union Politics at the University of Salzburg.

Dorte Sindbjerg Martinsen is Professor of Political Science at the University of Copenhagen.

ORCID

Michael Blauberger http://orcid.org/0000-0002-3581-400X

Dorte Sindbjerg Martinsen http://orcid.org/0000-0002-2113-0950

Additional information

Funding

The research for this article was part of a Norface Welfare State Futures (WSF)-funded project, TransJudFare [grant number 462-13-009].