National Responses to European Court Jurisprudence

Abstract

The power of the European Court of Justice (ECJ) to promote European integration through law has been broadly acknowledged, but the court’s domestic impact has received less attention and remains contested. In particular, the ambiguity of many ECJ judgments is said to have two opposed effects: According to one logic, legal ambiguity enables national policy-makers to contain the impact of court rulings, i.e. to ignore potentially broader policy implications. According to another logic, ambiguous case law provides opportunities for interested litigants to pressure national policy-makers into (anticipatory) adjustments. Which of these two logics prevails, it is argued, depends on the distribution of legal uncertainty costs between supporters and challengers of the regulatory status quo. The argument is supported by two in-depth case studies on the domestic responses to series of ECJ rulings concerning the free movement of capital (golden shares) and services (posted workers).

How do EU member states respond domestically to the rulings of the European Court of Justice (ECJ)? The court’s power to promote ‘integration through law’ has been broadly acknowledged and analysed. Less attention has been devoted to the ECJ’s role as an ‘agent of Europeanization’ (Panke 2007), i.e. to the actual policy impact of European jurisprudence at the domestic level. Disagreement exists regarding the extent and the mechanisms of ECJ-driven Europeanisation. A better understanding of ECJ-driven Europeanisation is crucial, however, for assessing whether indeed ‘the constitutional balance of the multilevel European polity … is upset by the “perpetual momentum” of ECJ case law’ (Scharpf 2012: 134).

Existing studies agree that ECJ rulings are often ambiguous and not self-implementing, but they reach very different conclusions regarding their domestic impact: according to one logic, legal ambiguity allows national policy-makers to follow ECJ rulings only in the concrete cases, but to ignore their broader policy impact (Conant 2002). According to another logic, ambiguous case law enables interested litigants to pressure national policy-makers into far-reaching adjustments (Schmidt 2008: 304). While both logics are theoretically plausible and supported by empirical evidence, the puzzle remains: what determines which logic prevails?

Domestic responses to ECJ rulings, it will be argued, depend on how the costs of legal uncertainty are distributed between supporters and challengers of the regulatory status quo (i.e. of existing domestic rules before their adjustment to ECJ jurisprudence). Time constraints, the population of similar cases and each party’s worst case scenario for an eventual ECJ ruling affect the distribution of uncertainty costs. If the challengers of the regulatory status quo have to carry the main burden of legal uncertainty, national policy-makers are likely to pursue a strategy of ‘contained compliance’. By contrast, if legal uncertainty is particularly costly for the supporters of the regulatory status quo, a strategy of ‘anticipatory obedience’ with ECJ jurisprudence is more likely.

The next two sections expose the puzzle of this article against the background of existing Europeanisation research and present the main theoretical argument. Subsequently, the argument’s plausibility is shown by two in-depth case studies on the domestic responses to series of ECJ rulings concerning the free movement of capital (golden shares cases) and the free movement of services (posted workers cases). Confronted with the court’s jurisprudence on Volkswagen and other golden shares cases, EU member states largely adopted a strategy of contained compliance, while the ECJ’s rulings on the regulation of posted workers in Rüffert and Laval triggered encompassing domestic reforms. This diversity of domestic responses to ECJ jurisprudence reflects different distributions of uncertainty costs between the parties involved in the legal conflicts.

The Domestic Impact of ECJ Jurisprudence

Literature on the ECJ and its role in promoting ‘integration through law’ abounds (Stone Sweet 2010). The same holds true for the issue of ‘Europeanisation’, i.e. the domestic impact of European integration (Börzel and Risse 2007). Yet relatively few studies focus on the intersection between these two broader research agendas, namely on the domestic impact of ECJ jurisprudence. On the one hand, studies on court-driven integration through law are mainly interested in novel legal interpretations established through case law, but they rarely include the successive process of political absorption and, thereby, tend to overestimate the court’s broader policy impact: ‘most accounts implicitly or explicitly assume that ECJ rulings are automatic catalysts for policy change and that innovative legal interpretation prompts wide-ranging reforms’ (Conant 2002: 15). On the other hand, studies on Europeanisation and member state compliance usually concentrate on the implementation of European secondary law and, therefore, tend to overlook domestic adjustments triggered by the court’s interpretation of European Treaty law. The constraining effect of ECJ-driven negative integration is particularly difficult to capture and, hence, domestic ‘non-decisions’ resulting from the court’s interpretation of European market freedoms are easily neglected (Töller 2010: 429f.).

Nevertheless, both literatures share two important insights concerning the domestic impact of ECJ jurisprudence. First, the interpretation of European Treaty law is often uncertain and ambiguous. Beforehand, ECJ rulings are difficult to predict, and afterwards their application remains contested. ECJ rulings do not prescribe specific policy responses and their implications beyond individual cases are often unclear. There is no doctrine of binding judicial precedent in EU law and core principles such as the proportionality test leave great room for interpretation when applied to new facts (Martinsen 2011: 945). Secondly, the court can hardly act as an ‘engine of integration’ or as an ‘agent of Europeanisation’ just by itself. Agency is required to provide the ECJ with cases in the first place and to turn ECJ judgments into real domestic change: ‘the Court’s case law may change legal norms, policy frames, and incentive structures for actors, but its jurisprudence is never self-implementing at the national level’ (Stone Sweet 2010: 33).

How far are legal ambiguity and the role of agency crucial for understanding domestic responses to ECJ jurisprudence? If we suppose that ECJ rulings are never self-implementing and, furthermore, exclude cases in which member state governments simply use court rulings as welcome pretexts for unpopular domestic reforms, we can derive two competing expectations regarding domestic responses to ECJ jurisprudence: According to the logic of contained compliance, most prominently represented by Lisa Conant, legal ambiguity provides loopholes for domestic policy-makers and allows them to avoid systematic reform: ‘Uncertainties that arise from indeterminate judicial principles dissuade general policy responses. As long as the application of legal principles remains open to debate, Member State officials generally maintain preferred policies’ (Conant 2002: 70). In this view, member states usually respond by implementing ambiguous ECJ rulings only with respect to the individual case treated in court, while disregarding more far-reaching implications (Conant 2002: 32). Governments from other member states which are not targeted by an ECJ judgment but pursue similar policies will typically deny any obligation and continue ‘business as usual’ (Conant 2002: 69). A domestic approach of contained compliance may be a strategic political decision, e.g. when member state legislatures deliberately adopt an attitude of ‘wait and see’ in the face of upcoming legal challenges (Slagter 2009: 191), or it may simply result from the unawareness of national administrations as regards potentially broader implications of ECJ case law (Conant 2002: 70).

Yet the ambiguity of ECJ case law might not always pay out in favour of member states. According to a different perspective, most pronouncedly taken by Susanne K. Schmidt, ambiguous ECJ jurisprudence provides an opportunity structure for interested litigants to indirectly pursue their domestic interest in the EU’s multi-level system. Enabled by uncertain case law, they can pressure domestic policy-makers into anticipatory obedience: ‘The higher the legal uncertainty arising from a Treaty rule and its interpretation, the more opportunities it offers for domestic actors to turn to the European courts in order to press for Europeanization’ (Schmidt 2008: 304). Regarded from that perspective, legal uncertainty is highly undesirable for domestic policy-makers. Political planning reliability is undermined if domestic policies are adopted or maintained only under the reservation that they might be challenged under European law at any time. ‘Contained compliance’ measures remain vulnerable to follow-up challenges and, thus, invite ever more judicial interference in domestic affairs (Scharpf 2012: 134). In contrast to political compromises, the decisions of independent courts are difficult to predict and impossible to control for national policy-makers. They may, therefore, prefer to compromise with potential litigants and anticipate possible legal challenges in designing domestic policies. In fact, such a strategy of anticipatory obedience may allow national regulatory autonomy to be preserved at least partially if it successfully forestalls ever more private litigation (Martinsen and Vrangbæk 2008: 182). Moreover, court rulings may create new legal uncertainty elsewhere and thus trigger adjustments even in member states which are not directly involved in individual ECJ cases (Schmidt 2008: 304).

In sum, member state governments can draw very different conclusions from the fact that ECJ jurisprudence often leaves considerable room for interpretation: they may follow a logic of contained compliance or one of anticipatory obedience. But what determines whether legal ambiguity is mainly perceived by national policy-makers as providing an escape route from strict compliance or as entailing a high risk of more and more challenging jurisprudence? Regarded from the perspective of potential litigants: when does ambiguous case law frustrate or encourage follow-up litigation?

The Costs of Legal Uncertainty

The answer to the puzzle just outlined, it will be argued, rests on the following three claims: legal uncertainty is costly; the distribution of uncertainty costs varies; and depending on whether challengers or supporters of the regulatory status quo carry the main share of uncertainty costs, contained compliance or anticipatory obedience are more likely.

To begin with, legal uncertainty, understood as a lack of predictability of law (Schmidt 2008: 300), is inefficient in a rule of law system and, therefore, costly for those affected. Hence, if legal uncertainty persists despite the court’s jurisprudence and unless the parties to a conflict reach a stable compromise by themselves, they have to carry uncertainty costs.

These uncertainty costs, however, can be distributed differently between supporters of the regulatory status quo, i.e. governments and the beneficiaries of existing regulation, and its challengers. Three main factors determine the cost distribution: time constraints, the population of similar cases, and the parties’ worst case scenario for an eventual ECJ ruling.

Time constraints: legal conflicts which are decided by the ECJ take time – about 16 to 20 months on average from the submission of an infringement action by the Commission or a preliminary reference from a national court to the final ECJ ruling (European Court of Justice 2011: 96). Moreover, the referral to the ECJ comes only at a very late stage in the overall proceedings. While the Commission does not publish information on the average duration of its infringement actions, the list of on-going proceedings includes cases from the 1990s (European Commission 2011a: 260, 273). Knowing about the potential length of legal proceedings, national policy-makers and potential litigants may have a strong interest in avoiding sustained legal uncertainty. On the one hand, if an ECJ judgment challenges an active domestic policy, member state governments are under particular time pressure to re-establish a firm legal basis for their action. Even if elected politicians themselves might not care too much about potential legal conflicts in the future, national administrations and potential beneficiaries of domestic policies have an immediate interest in avoiding legal uncertainty, e.g. when allocating public tenders or receiving subsidies (Blauberger 2009: 1041). On the other hand, an economic actor or citizen acutely suffering from a restriction to the free movement, e.g. missing out some vital cross-border business, can hardly count on a judicial solution in the short or medium term. Thus, persistent legal uncertainty gets costlier the tighter one’s time constraints are.

Population of similar cases: individual court rulings are always more specific than general legislation, but they can nevertheless differ with regard to the generalisability of the legal issues addressed and, thus, potentially target many or few similar legal situations. Moreover, the expected benefits from a successful legal challenge may motivate more or less interested litigants. Access to the judicial system can be rather permissive, e.g. when secondary legislation establishes clear rights and obligations, or restrictive, e.g. when private parties largely depend on infringement action on the part of the Commission (Conant 2002: 53, 73). As a result, member state governments can become confronted with few or many follow-up legal conflicts after an ECJ judgment. The larger the population of similar cases, the costlier it becomes for national policy-makers to prevent legal challenges through political compromises and to defend existing regulation in the courts. By contrast, if interested litigants cannot count on potential allies, they have to bear the entire costs of upholding the threat of a legal challenge and, eventually, of sustaining the case in court by themselves.

Worst case scenario: implicitly or explicitly, parties to a legal conflict develop worst case scenarios for an eventual ECJ ruling. For the challengers of the regulatory status quo, this scenario is relatively easy to establish: it consists of a judicial approval of the regulatory status quo. For the supporters of the regulatory status quo, however, the potential implications of a negative ECJ ruling are often highly uncertain in advance and may only materialise after several rounds in court. Thus, even tangible costs such as infringement penalties enter the strategic calculus of the supporters as an uncertain and subjective element. Depending on the legal framework of a particular policy, ECJ rulings may involve different kinds of financial risks, e.g. recovery decisions in the field of state aid, contractual penalties in public procurement, state liability and even private liability for infringements of EU law. Establishing liability, for instance, is difficult and far from coherent in practice (Conant 2002: 62), but already the possibility of such a worst case significantly alters the uncertainty costs for national policy-makers. In harmless cases, member state governments can afford to wait for judicial defeats, even though ‘previous ECJ case law virtually guarantees they will lose’ (Conant 2002: 72). A disputed individual claim to social benefits, for example, can still be satisfied after an ECJ ruling. If such a ruling risks motivating many follow-up claims, however, national policy-makers may prefer to generously satisfy the individual claim in order to avoid a binding ruling (Obermaier 2008: 751).

Taken together, these factors determine the distribution of uncertainty costs between supporters and challengers of the regulatory status quo and, ultimately, the kind of domestic response to the ECJ’s jurisprudence to be expected: The higher the burden of legal uncertainty for the supporters of the regulatory status quo, the more likely they are to adopt a strategy of anticipatory obedience in order to avoid sustained legal conflicts. By contrast, the higher the uncertainty costs of the challengers, the easier it becomes for national policy-makers to resist adjustment pressures and the more likely they are to adopt a strategy of contained compliance.

Case Studies

The remainder of this text assesses the plausibility of the theoretical argument just outlined by comparing member states’ responses to two series of ECJ judgments on the free movement of capital (golden shares) and services (posted workers). Starting from series of judgments rather than isolated cases allows a better understanding of member state responses in the context of existing case law. Moreover, the analysis not only investigates those member states directly targeted by the court’s rulings, but also looks for indirect effects in other member states.

The two series of judgments have been chosen because their legal and political background was largely similar, but the uncertainty costs were distributed very differently. In legal terms, both series affected policies which are not harmonised at EU level, but which are, according to the ECJ, constrained by European market freedoms. They are, thus, typical examples of ECJ-driven negative integration. In terms of political salience, Volkswagen¹ and Laval² were certainly among those ECJ rulings which received the highest public attention in recent years. The court was strongly criticised by member states’ governments for intruding into essentially domestic policies. At the same time, liberal economic interests such as foreign investors and service providers welcomed the judgments for their anti-protectionist thrust. As the judgments still left considerable room for interpretation, both kinds of responses outlined above were prima facie plausible: domestic policy-makers trying to contain compliance in order to preserve regulatory autonomy, and private actors threatening to engage in follow-up litigation in order to push for encompassing reform.

As will be shown, a strategy of contained compliance prevailed in the aftermath of Volkswagen and other golden shares cases, while the policy responses to Laval and Rüffert went further and even tried to anticipate future legal challenges.

Volkswagen and the Golden Shares Cases

The ECJ’s Volkswagen ruling is usually regarded as being part of a larger series of ‘golden shares’ judgments by the Court since 2000 (Zumbansen and Saam 2007: 1028). All of these cases concerned public special rights in formerly state-owned enterprises, e.g. regarding the appointment of board members or voting privileges in important company decisions. The special rights are either based on the public ownership of a special share in a private company (a golden share in a narrow sense), or they are established by regulation (Grundmann and Möslein 2004: 2).

Member state governments usually justify golden shares as a necessary means to secure the provision of services of public interest in sensitive sectors such as energy, telecommunications or banking. By contrast, the ECJ adopted a restrictive position towards golden shares which are often suspected to protect ‘national champions’ from foreign influence rather than serving the provision of services of public interest: it interpreted public special rights broadly as restrictions to the free movement of capital and, rejecting the justifications forwarded by member state governments, it largely considered the restrictions disproportionate. Yet the court has not entirely ruled out the possibility of golden shares which are compatible with EU law and has left some room for legal interpretation. In 2002, it permitted golden shares in the Belgian energy sector against the objections of the European Commission.³

The German response to Volkswagen. The ECJ delivered its judgment on Volkswagen on 23 October 2007 and supported the claim of the Commission that Germany’s Volkswagen law constituted an unjustified restriction of the free movement of capital. The Commission had challenged three provisions of the Volkswagen law: the public privilege to appoint two members of Volkswagen’s supervisory board; the cap of voting rights at 20 per cent for any shareholder (including those holding more than 20 per cent of Volkswagen shares); and the fixing of a specific blocking minority of 20 per cent of the general assembly of shareholders (whereas German law on public limited companies fixes the blocking majority at 25 per cent). Although the latter provisions applied to private and public shareholders alike, they were seen by the court as de facto favouring the Land of Lower Saxony, which holds precisely 20.01 per cent of Volkswagen shares.

The German response to Volkswagen marks a striking attempt to contain compliance with the Court’s ruling. A revised Volkswagen law was adopted in December 2008 which removed only the first two contested provisions, but kept the 20 per cent blocking minority. Due to an ambiguity in the ECJ’s ruling, Germany could argue that the fixing of the blocking minority had not been challenged independently, but only ‘in conjunction with’ (no. 83(1) of the judgment) the voting cap (Gerner-Beuerle 2012: 102). In addition to the revised Volkswagen law, Lower Saxony together with the other major shareholders of Volkswagen reintroduced parts of the contested special rights into the company’s articles of association in December 2009, including the right of Lower Saxony to appoint two members of Volkswagen’s supervisory board and the 20 per cent blocking minority (Gerner-Beuerle 2012: 104). Again, this measure took advantage of persistent legal ambiguity, namely regarding the question whether Volkswagen’s articles of association, autonomously adopted by its shareholders, can infringe European Treaty freedoms. Dissatisfied with Germany’s response, the Commission ultimately referred the case back to the ECJ.⁴

So far, legal uncertainty allowed German authorities to contain compliance in the Volkswagen case. First and foremost, the German government was able to play for time. The Commission’s first informal attempts to tackle the Volkswagen law date back to 2001; the infringement proceeding was initiated by a formal notice to the German government in March 2003; more than four years elapsed after the court’s first ruling before the Commission referred the case back to the ECJ. After the judgment, the Commission warned German authorities very early on that it would take a close look at the revised law, but a strategy of anticipatory obedience with the Commission’s claims was explicitly rejected by Germany’s minister of justice and no further adjustments were introduced into the law after December 2008.⁵ By contrast, time (or the lack thereof) was crucial for one of the major interests behind the Commission’s infringement proceeding, namely for Porsche and its planned takeover of Volkswagen. Since 2005, Porsche had secretly prepared a takeover of Volkswagen, seeking to incrementally acquire 75 per cent of Volkswagen shares and counting on an abolition of the Volkswagen law. When the financial crisis hit in 2008 and the revision of the Volkswagen law did not remove the 20 per cent blocking minority, however, the plan collapsed (Gerner-Beuerle 2012: 101–2).

In addition, the peculiarity of the Volkswagen law and the fact that German authorities had little to fear even from a negative ECJ judgment allowed them to contain compliance. Prior to the ECJ’s ruling in 2007, Germany argued that Volkswagen was historically and legally completely different from the golden shares cases and, hence, it was not affected by the ECJ’s case law on that matter.⁶ Afterwards, the relevance of the golden shares jurisprudence could no longer be generally denied, but the ambiguities of the Volkswagen judgment described above still allowed the narrow revision to be defended with very case-specific arguments. The population of similar cases, thus, was zero. Given the uniqueness of the Volkswagen law, German authorities never had to fear any similar challenges elsewhere. At the same time, the Commission had to invest considerable efforts to purse the infringement proceedings against Germany, given its limited enforcement capacities and the specifics of the case. Apart from an unsuccessful domestic legal challenge against the 20 per cent blocking minority in Volkswagen’s articles of association,⁷ Porsche largely depended on the Commission’s determination to continue its infringement action, but when the Commission remained inactive until late 2011, Porsche’s planned takeover had already vanished.

Finally, Germany did not suffer any sanctions from the court’s ruling in 2007 and the Commission’s on-going infringement proceedings. In the worst case of a second negative judgment, Germany would face financial penalties, but even these could be limited by another effort to contain compliance, i.e. by quickly removing the blocking minority from the Volkswagen law while keeping it in the articles of association (Holle 2010: 19–20).

Similar responses in other golden shares cases. Volkswagen probably was an extraordinary case in terms of public attention, but the pattern of contained and delayed compliance just analysed is paradigmatic for many golden shares cases. Typically, many member states do not give in to the Commission’s demands unless the court finally finds an infringement of European Treaty rules and even after the ECJ has ruled, a considerable number of conflicts persist. Out of 26 proceedings concerning golden shares so far, the court had to rule on 15 cases and only one case was decided positively by the court (see note 4). Compared to member states’ average compliance record, these figures are striking. According to the Commission, 88 per cent of all infringement proceedings are closed at earlier stages, i.e. without any involvement of the ECJ (European Commission 2011a: 3–4).

As in the case of Volkswagen, member states’ recalcitrance to compromise with the Commission and to draw lessons from the increasingly rich ECJ case law on golden shares can be explained by the distribution of uncertainty costs. To begin with, the hardest-fought conflicts about golden shares revolve around foreign investments in companies which are protected by public special rights and time-consuming conflicts frustrate investors. For example, when the Austrian oil and gas company OMV publicly declared its intention to merge with its Hungarian competitor MOL in September 2007, the Hungarian parliament adopted the Act CXVI/2007 ‘relating to companies of particular importance for the security of public services’ (Butler 2011: 637) in October 2007. This law, unofficially labelled ‘Lex MOL’, aimed at preventing what was perceived as a hostile takeover of MOL by a foreign enterprise under Austrian state control and with opaque relations to Russian Gazprom.⁸ Eventually, OMV withdrew its bid for MOL when, somewhat schizophrenically, the European Commission criticised Hungary’s infringement of the internal market, but also voiced concerns regarding the competition aspect of the planned merger (Butler 2011: 642). In a similar case from the energy sector, the Spanish government prevented the takeover of Endesa by the German enterprise Eon. Spain first played for time by blocking the merger and when Spanish Acciona and Italian Enel announced a joint offer for Endesa, Eon withdrew its bid (Harker 2007: 517). In another legally complex case, the prospect of a lengthy conflict convinced the Italian bank Unicredito to give in to the Polish government before the Commission could even push its infringement proceeding further. The Commission had already approved the merger of Unicredito and German HVB in 2005, but the Polish government feared that the merged bank would hold a dominant position on the Polish market. Therefore, claiming special rights from a privatisation agreement with Unicredito from 1999 concerning the former state-owned Polish bank Pekao, Poland required Unicredito to sell a large share of the Polish HVB branches to a third party. In addition, Poland filed a complaint against the Commission’s merger decision to the Court of First Instance (CFI). In order to avoid legal uncertainty, Unicredito agreed to the sale of its Polish HVB branches (Farmer 2008: 199–201).

Apart from the importance of time constraints, more commonalities between the responses to Volkswagen and other golden shares cases exist. The population of similar cases is always very small or zero. To deny the applicability of previous golden shares jurisprudence, member state governments typically emphasise the uniqueness of their own golden shares and try to benefit from remaining legal gaps. For example, Hungary framed its ‘Lex MOL’ in general terms of securing energy supply and did not mention any specific enterprise in order to distinguish itself from Germany’s Volkswagen law (Butler 2011: 638). As an alternative strategy, member states explicitly refer to the only permissive judgment of the ECJ regarding Belgian golden shares and argue that their golden shares legislation was drafted according to the Belgian model (Adamczyk and Barański 2010: 96). In any event, member states interpret existing case law on golden shares so as to largely dispute potential legal constraints for their own golden shares. As a consequence, challengers of existing golden shares need to tackle each case individually. Moreover, private actors have to rely on the Commission which has only limited enforcement capacities and which carries the main burden of proof when referring infringement cases to the ECJ.

Moreover, as with Volkswagen, member state governments often do not appear to be deterred by the worst case of a negative ECJ judgment. The Portuguese government already lost four golden shares cases in court, the Italian government lost three, but nevertheless they fought these cases relentlessly. Interestingly, the stakes have risen due to the sovereign debt crisis. The worst case scenarios for the Portuguese and Italian governments have become much more severe. The Commission managed to link the issue of golden shares to the negotiations on the Portuguese rescue package and to Italian reforms under the Monti government. As a consequence, Portugal abolished its golden shares (European Commission 2011b: 26) and Italy avoided a new ECJ referral by reform of its golden shares legislation.⁹ By and large, however, member states’ responses to the ECJ’s golden shares jurisprudence provide compelling evidence for a logic of contained compliance.

Laval and the Posted Workers Cases

In late 2007 and early 2008, the ECJ decided a series of cases on the applicability of domestic labour law to workers posted from other member states: Laval, Rüffert¹⁰ and Commission v Luxembourg.¹¹ In all three cases, the ECJ found national provisions to restrict the free movement of services under European Treaty law as specified in the Posted Workers Directive 96/71/EC. According to the judgments, member states were only allowed to subject posted workers to a ‘nucleus of mandatory rules for minimum protection’, but not to their entire regulation applicable to domestic workers, in particular not to the terms and conditions of non-universally applicable collective agreements.

At the same time, the Court explicitly acknowledged member states’ room for manoeuvre by stating that ‘the material content of those mandatory rules for minimum protection … may accordingly be freely defined by the Member States, in compliance with the Treaty and the general principles of Community law’ (No. 60 of the Laval judgment) and, therefore, member states were ‘free to choose a system … which is not expressly mentioned in that directive (on Posted Workers), provided that it does not hinder the provision of services between the Member States’ (No. 68). Obviously, these extracts from the court’s rulings are highly ambiguous, e.g. regarding the scope of ‘minimum protection’ or the reference to systems ‘not expressly mentioned’ in the directive and, hence, the judgments’ implementation was contested at the domestic level.

Swedish and Danish responses to Laval. The Laval judgment of the ECJ responded to a preliminary reference from the Swedish Labour Court and marked the climax of a conflict between the Swedish building workers’ union (Byggnads) and the Latvian construction company Laval un Partneri. Laval wanted to post Latvian workers in order to build a school in Sweden, but instead of reaching an agreement with Swedish trade unionists, the company signed a collective agreement with a Latvian trade union. When Swedish trade unionists denied the validity of this agreement and took collective action by blocking the construction site, Laval took the conflict to court. Eventually, the ECJ acknowledged the fundamental right of trade unions to take collective action, but found the specific actions in this case, blocking the construction site, disproportionate. In particular, the ECJ criticised the lack of precise provisions on the determination of minimum pay, given that no statutory minimum wage exists in Sweden and collective agreements are not universally binding. In addition, the Swedish ‘Lex Britannia’, which permitted collective action irrespective of existing collective agreements under foreign law, was found to be discriminatory.

In response to the Laval judgment, the Swedish government and social partners undertook a comprehensive reform of the regulation of posted workers which went far beyond the individual case and also aimed at anticipating future conflicts with EU law. An inquiry was mandated to explore possible responses and legislation was adopted in November 2009 which abolished the contested provision of the Lex Britannia and specified conditions for lawful collective action under the Posting of Workers Act. According to these conditions, collective action against a foreign employer is only permitted if it aims at establishing working conditions and wages which correspond to those concluded at the central level, which are applied throughout Sweden to corresponding workers in the same sector, and which remain within the nucleus circumscribed by the Posted Workers Directive (Rönnmar 2010: 285). As a consequence, Swedish trade unions had to translate their normal collective agreements into new ‘posting agreements’, which are restricted to the nucleus of minimum conditions and supposed to increase transparency for foreign employers with regard to the calculation of minimum wages.

Notably, the Laval judgment was not only followed by legislative reform in Sweden, but in Denmark as well. Denmark was not directly targeted by the ECJ’s judgment, but given broad similarities between the Swedish and Danish models of regulating posted work, the Danish government was concerned about the potential implications of the ECJ’s ruling in Denmark. Therefore, the Danish government and social partners engaged in a similar reform process which was concluded one year earlier than the Swedish reform (Malmberg 2010). The Danish reform differs from the Swedish one in important respects, e.g. regarding the broader definition of minimum wages, but – as in the Swedish case – it aims at avoiding and anticipating legal conflicts arising in the context of EU Treaty law (Blauberger 2012: 117).

How can we explain these domestic responses which clearly went beyond the individual case in Sweden and even involved legislative reform in Denmark? In contrast to Volkswagen and the golden shares cases, legal uncertainty posed a major problem for member states’ governments and, even more so, for trade unions as the intended beneficiaries of existing regulation. Time pressure to overcome legal uncertainty was particularly strong in Denmark; the population of similar cases was high. At the time of the Laval ruling, Danish trade unions were involved in a considerable number of collective action measures against foreign service providers. As long as the legal basis of these measures was not firmly re-established in the light of Laval, the threat of legal challenges under EU law was urgent. The possibility of a ‘Danish Laval’ case had already become apparent in 2005, but the case was solved domestically without reference to the ECJ and based on a permissive interpretation of EU Treaty law (Neergaard and Nielsen 2010: 485). After Laval, such a legal interpretation was hardly sustainable and, therefore, the Danish reform was a proactive adjustment in order to ensure the continued lawfulness of collective actions measures. In Sweden, time pressure was less pronounced as no other collective action measures comparable to Laval were under way. Still, strikes and sympathy actions are used regularly by Swedish trade unions to enforce their demands in collective bargaining and, thus, their legal basis needed to be revised as well. Under the shadow of the Laval judgment and its pending domestic implementation, Swedish trade unions deplored a considerable ‘chilling effect’ (Rönnmar 2010: 286), i.e. trade unions were more cautious in their use of allegedly unlawful collective action measures and, thus, they were weakened in their collective bargaining position.

Finally, Laval fundamentally changed the worst case scenario for Swedish trade unions. In the run-up to the ECJ’s ruling, Swedish trade unionists were rather confident that they would win the case and took an uncompromising position (Davesne 2009: 12). Their worst case scenario was an ECJ ruling which would have obliged the referring national court to examine the proportionality of collective action measures in the light of European law. Unexpectedly, however, the ECJ decided the issue itself and declared the measures against Laval disproportionate. Even worse for Swedish trade unions, the conflict continued at the domestic level as Laval claimed financial damages for its losses as well as punitive damages for the breach of European law. The Swedish Labour Court partly affirmed the liability of Swedish trade unions in December 2009, ordering them to pay about €55,000 of punitive damages (Rönnmar 2010: 281). In sum, Swedish trade unions underestimated the possible outcome of the Laval conflict, which partly explains their uncompromising attitude during the conflict. Afterwards, however, legislative reform and a more cautious approach towards collective action measures against foreign service providers aimed at avoiding similar legal conflicts in the future, not the least because of the threat of punitive damages. Avoiding follow-up conflicts to Laval was also seen as essential for the legitimacy of Swedish and Danish autonomous collective bargaining which rests on the self-image of social partners being able to ‘crack the nuts ourselves’ (Malmberg 2010: 13).

German Länder responses to Rüffert. A similar pattern of broad domestic reform triggered by the ECJ’s jurisprudence on posted workers can be observed in the Rüffert case. The conflict concerned the Public Procurement Act of the German Land Lower Saxony, in particular the requirement for companies to commit themselves to local collective agreements in order to be eligible for public tenders. According to the ECJ’s ruling of 3 April 2008, Lower Saxony’s Public Procurement Act imposed an unjustified restriction on the freedom to provide services under the Posted Workers Directive, in particular as it protected only workers in the public sector and referred to collective agreements that were not universally applicable.

The Rüffert judgment led to a response not just in Lower Saxony, but it triggered a process of legislative reform throughout almost all German Länder. Immediately after the ECJ’s ruling, all Länder with similar legislation as in Lower Saxony disapplied the contested provisions in April and May 2008. Subsequently, from 2008 to 2011, 13 of 16 Länder revised or started to revise their public procurement legislation: three simply removed the contested provisions; six inserted new, EU-compatible provisions on social criteria into their public procurement legislation; and four are still reforming their legislation (Blauberger 2012: 120). The three Länder which remained inactive did not have any provisions similar to those in Lower Saxony when the Rüffert case was decided. In terms of material content, the reforms varied considerably among German Länder depending on the party composition of their governments (Sack 2012), but regardless of these different party positions all Länder enacted encompassing reforms rather than remaining inactive or trying to contain compliance.

As with Laval, inaction after Rüffert would have implied significant legal uncertainty which was particularly problematic for public authorities. The immediate suspension of the contested provision in Lower Saxony and similar provisions in all other Länder within less than two months shows the urgency of the reaction. In fact, representatives of the political Left and trade unionists tried to advocate a strategy of contained compliance directly after Rüffert by hinting at the specifics of the case and by negating its broader implications, e.g. for the legislation of German Länder other than Lower Saxony (Hänlein et al. 2008: 23–24). Such a strategy, however, was clearly rejected by all Länder governments with reference to the legal uncertainty it would have implied for a great number of similar public tenders. In the direct aftermath of Rüffert, public authorities reported an increase of private inquiries concerning recent or on-going public tenders (Hänlein et al. 2008: 85). Ensuring legal certainty became the dominant issue for numerous expert surveys issued by Länder governments in the process of legislative reform (Blauberger 2012: 199).

Finally, legal uncertainty was not just unattractive for public authorities because of the great population of similar cases, but because of the significant financial risk due to liability suits. After the ECJ had ruled, the Rüffert conflict itself was settled confidentially out of court, but follow-up conflicts would have clearly implied a high risk of financial liability. In one of the earliest legal surveys after Rüffert, commissioned by the Land Bremen, the option of ‘not buckling’ under the court’s jurisprudence and deliberately taking the risk of ‘an unsuccessful applicant claiming punitive damages’ is explicitly addressed (Däubler 2008: 12, own translation). This worst case was to be averted through the suspension and revision of public procurement legislation.

In sum, the ECJ’s Laval and Rüffert judgments triggered processes of encompassing reform, not just limited to the individual cases and not restricted to the governments directly involved in the conflicts. The absence of follow-up ECJ cases to the posted workers conflicts is telling in itself: at first sight, it might seem paradoxical given the great population of potentially similar conflicts regarding posted workers compared to the relatively few instances of golden shares. Yet member states have reacted to the posted workers cases in a more anticipatory and encompassing way precisely to avoid endless conflicts and the corresponding legal uncertainty.

Conclusion

This article started out from the legal ambiguity inherent in many ECJ rulings and asked for its implications when these rulings are applied domestically. Two opposed expectations were derived from existing Europeanisation research: member state governments may opt for a strategy of contained compliance, i.e. they use loopholes in ECJ rulings to minimise their domestic impact, or they can adopt a strategy of anticipatory obedience, i.e. they engage in encompassing reforms to reduce the pressure from interested litigants. Both logics were found to be highly plausible, but underspecified: Which logic prevails, it was argued, depends on the distribution of uncertainty costs between supporters and challengers of the regulatory status quo.

As an empirical illustration, two series of cases with very different distributions of uncertainty costs, but otherwise broad similarities, were selected. Given these similarities, other potential explanations for the diverse national responses to the posted workers and golden shares jurisprudence can be largely excluded. Both series of rulings received great public attention as they touched upon politically highly salient issues (see Blauberger 2012: 109, Gerner-Beuerle 2012: 100): collective agreements and public prerogatives in strategic sectors. Political salience, thus, cannot account for the observed variation and the different domestic responses also do not correspond to party-political cleavages. By and large, opposition to the court’s jurisprudence stretched across all major parties in the posted workers as well as the golden shares cases, yet the responses differed. In Laval, for instance, even the conservative Swedish prime minister Reinfeldt had sided with trade unions (Davesne 2009: 13). Rüffert was only welcomed by the small German liberal party (FDP), which abolished the contested provisions when it was part of the governing coalition (Sack 2012), but even those social democrat Länder governments that were not directly targeted by the ruling engaged in encompassing reforms. Most strikingly, the government of Lower Saxony was directly involved in the Rüffert and Volkswagen conflicts, but legislative responses differed greatly. The examples from Lower Saxony and the consistent pattern of contained compliance regarding golden shares across member states are also at odds with the argument about general ‘worlds of compliance’ in the EU (Falkner et al. 2005). Still, the Swedish and Danish responses to the posted workers cases fit the arguments that these member states are particularly ‘law observant’ (Falkner et al. 2005: 331) and sceptical towards settling political conflicts in court (Wind 2010: 1047).

A cautionary remark, therefore, is in order. Research on Europeanisation through case law is still surprisingly scarce and could greatly benefit from further studies – not only testing the argument forwarded above and alternative accounts against a broader set of cases, but also including other possible responses to European jurisprudence (e.g. when national policy-makers use the ECJ merely as a scapegoat for unpopular reforms) and assessing the relative weight of different responses. The overall share of instances of contained compliance or anticipatory obedience cannot be established on the basis of a few case studies, but the present study has shown the existence of these two distinct logics and has identified factors which make them likely to apply: when member state governments are able to play for time, when they do not face the risk of financial sanctions in the short or medium term, and when challengers have to fight for their cause individually, a logic of contained compliance is likely to prevail (as in the golden shares cases). By contrast, when legal uncertainty undermines political planning capacity or even involves great financial risks, and when the spectrum of potential litigants becomes too vast, member state governments anticipate future legal challenges by adjusting domestic regulation (as in the posted workers cases).

Acknowledgements

I am grateful to Martin Höpner, Rike Krämer, Tilman Krüger, Fritz W. Scharpf, Bernd Schlipphak, Daniel Seikel and two anonymous reviewers for their comments. I benefited particularly from my collaboration with Susanne K. Schmidt and her project A6 at the Collaborative Research Centre 597 at the University of Bremen.

Notes

1. Case C-112/05, Commission v. Germany (Volkswagen), ECR 2007, I-8995.

2. Case C-341/05, Laval un Partneri Ltd, ECR 2007, I-11767.

3. Case C-503/99 Commission v. Belgium, ECR 2002, I-4809.

4. Case C-94/12, Commission v. Germany (Volkswagen), not yet decided.

5. ‘Vorauseilender Gehorsam? Abgelehnt!’ (‘Anticipatory Obedience? Refused!’, own translation), Sueddeutsche Zeitung, 19 September 2008.

6. See no. 42 and no. 95 of the opinion of the Advocate General Colomer on the Volkswagen case, delivered on 13 February 2007.

7. Landgericht Hannover, judgment of 27 November 2008, ZIP 2009, 666.

8. ‘Ungarn peitscht “Lex MOL” durch, setzt auf Zeit’ (‘Hungary pushes through “Lex MOL” and plays for time’, own translation), Die Presse, 6 October 2007.

9. ‘Italy limits “golden share” rules to avoid hefty EU fines’, EUbusiness, 9 March 2012.

10. Case C-346/06 Rüffert, ECR 2008, I-1989.

11. Case C-319/06, Commission v. Luxembourg, ECR 2008, I-4323. The court’s judgment left practically no room for interpretation regarding the contested provisions in Luxembourg’s regulation of posted workers and, hence, did not allow for a strategy of contained compliance or require any anticipatory obedience.