The European Court of Justice and its political impact

Abstract This article reviews recent advances in the study of the European Court of Justice (ECJ) and its political impact at the European and member state levels. New quantitative as well as qualitative analyses show with great empirical precision that member state preferences guide the Court. The article summarises these findings, but argues that greater attention needs to be given to the (over-)constitutionalisation of EU law in order to fully capture the political impact of ECJ jurisprudence. Even if European judges are less activist than is often assumed and individual decisions are more restrained in the face of member state opposition, incrementally, case law evolves in a highly expansive fashion. And, exercising caution regarding unrealistic expectations about quasi-deterministic judicial law-making, it is found that the Court’s constitutionalised jurisprudence impacts heavily on European and member state policy-making.

The European Court of Justice (ECJ) has long been hailed as an independent motor of European integration. Yet recent work has shown with great empirical precision how much the Court is conditioned by member state preferences. Quantitative studies have analysed the influence of member state observations on the Court's jurisprudence (Carrubba and Gabel 2015;Carrubba et al. 2008;, and qualitative case studies have traced political corrections in response to ECJ case law (Martinsen 2015a). Other leading scholars have questioned the empirical validity of these claims (Davies 2016;Stone Sweet and Brunell 2012). In this review essay, we enquire into the theoretical significance of recent empirical work on the ECJ. What are the implications for our understanding of the political system of the EU if the Court can be shown to pay due regard to the preferences of the member states?
The long debate between supranationalism and intergovernmentalism gives us a misleading answer. A court paying tribute to member state preferences, we claim, is not necessarily a court having little impact. This answer would CONTACT Michael Blauberger michael.blauberger@sbg.ac.at OPEN ACCESS overlook the roots of the Court's impact: the direct effect and supremacy of the EU Treaties. These doctrines, established in the 1960s, effectively transformed the European Economic Community (EEC) Treaty into a constitution, in all but name. But an intergovernmental treaty makes for a very special constitutional framework. While core elements of traditional constitutions, such as rule of law principles and fundamental rights, remain underdeveloped, the EU Treaties are replete with detailed policy goals. The former German constitutional judge Dieter Grimm has coined the term 'over-constitutionalisation' for the unusual nature of the EU's Treaties (Grimm 2016). 1 We argue, in line with his analysis, that only by reflecting 'over-constitutionalisation' can we understand the broader impact of the Court. Interpreting a treaty with detailed policy prescriptions, the Court's rulings have direct implications for policy-making at the European and domestic levels. And due to the unanimity rule for treaty changes, overruling the ECJ is even more difficult than in the context of national constitutional jurisprudence. This prevalence of non-majoritarian decision-making in the EU is emphasised by Grimm's analysis.
In the following, we summarise the recent work on the Court and ask what over-constitutionalisation means for the political system of the EU and its member states. We start by discussing studies that show a considerable congruence between the Court's rulings and member state preferences. Such a narrow focus on Court restraints in individual disputes, however, fails to capture the expansive evolution of case law over time. We then turn to the few studies that systematically analyse the Court's impact on legislative politics, and on this basis emphasise the importance of over-constitutionalisation.

Judicial restraint vs expansive case law
Recent quantitative studies have found significant effects of member state political signals on ECJ case law (Carrubba and Gabel 2015;. Carrubba and Gabel mainly base their study on the summaries of member state interventions that were published until 1994 (Carrubba and Gabel 2015: 70). Larsson et al. received access via the Swedish Foreign Ministry to more recent, unpublished summary reports of member state written observations and compiled a dataset for the period between 1997 and 2008 (Larsson and Naurin 2016: 392). Both groups of authors coded member state positions and ECJ rulings in order to assess to what extent the Court follows or violates member state preferences.
The two studies analyse distinct mechanisms: legislative override, which requires collective action at the EU level , and non-compliance, which results from unilateral action at the domestic level (Carrubba and Gabel 2015). The underlying logic, however, is very similar: by submitting written observations to ongoing cases, member state governments threaten countervailing measures at the European or national levels against unwelcome rulings. Larsson and Naurin (2016: 382) argue that judges are influenced by the perceived risk of legislative override. And Carrubba and Gabel (2015: 45) hold that the Court only tends to interpret EU law expansively if the domestic costs of compliance are not getting too high and member state governments do not threaten to infringe European rules. In sum, according to the findings of these authors, the Court is more restrained than previously acknowledged when faced with strong member state opposition.
Apart from the Court's limited ability to engage in judicial activism, Larsson and Naurin also find little evidence for any political bias of ECJ judges that would go against member state preferences. Unsurprisingly, the Court and the Commission are found to be systematically more pro-European than member states -in particular, when EU governments disagree on political-economic grounds (Larsson and Naurin 2015: 24f.). However, ECJ judges do not appear to lean towards any particular liberal or social welfare model or variety of capitalism, but rather position themselves in between these different models (Larsson and Naurin 2015: 10-27).
What does this most impressive empirical work imply for the importance of the Court and its case law? If its rulings reflect intergovernmental preferences, if 'everything happens conditionally on implicit government acquiescence' (Carrubba and Gabel 2015: 213), can we still call the Court a motor of integration? In her recent book, Karen Alter (2014: 338) argues that it 'would be surprising indeed' if 'judges systematically and generally ignore[d] government opinion' . And yet, even in the light of the empirical data discussed above, we should not entirely dismiss the possibility of ECJ activism that goes beyond governments' preferences (i). Even more importantly, Court activism is just one way towards 'integration through law' . This is also one of the main findings of an edited volume of eminent legal scholars, namely that 'understanding judicial activism involves looking well beyond the decisions of the European Court themselves and into the very foundations of the EU' (Muir et al. 2013: 9). Rather than focusing narrowly on the Court and individual activist judgments, we therefore argue that greater attention has to be given to the ways in which the Court's case law expands incrementally over time (ii) and assumes relevance erga omnes (iii).
(i) Can interventions of governments be assumed to truly reflect their preferences? Carrubba and Gabel as well as Larsson and Naurin assume that governments rationally intervene in all those cases that matter to them. However, case studies show that this is not necessarily true, and governments do not always intervene, although rulings affect them (Obermaier 2008: 23-5). Possibly they are not aware of the implications of court cases arising in other member states. It is difficult for governments to assess the consequences of all possible turns that a ruling could take, as it is to assess how a ruling could favour or hamper future policy choices. Or governments may hope that their infraction of EU law will not come to light if they keep quiet. Interventions into cases thus mirror political salience only incompletely. In addition, preferences of governments are fluid, not only when accounting for governmental changes. Within coalition governments, there are frequently differing policy preferences. As the EU often serves purposes of blame avoidance, not all rulings which imply shifts in domestic policy are necessarily unwelcome. But even if governments concur in court-driven changes, this does not necessarily imply that these changes could have found domestic majorities without the Court. Government briefs, thus, do not simply mirror the extent of accepted domestic policy change. In sum, the Court may rule against incumbent governments to a greater extent than is acknowledged in existing studies based on member state written observations.
(ii) How are Court rulings linked to policy change? Transformative change need not be abrupt, but may evolve incrementally (Streeck and Thelen 2005: 9). In this respect, the Court's high case load allows for many small steps. Hence, even if the ECJ balances all claims cautiously and, by and large, respects strong member state objections in its rulings, its case law may develop a highly expansive effect over time. To capture this expansive effect, we need to analyse lines of jurisprudence rather than individual rulings. Moreover, it is necessary to account for the qualitative importance of legal issues. The Court may back down on a high number of relatively insignificant issues, while significant changes might result from the minority of issues in which the Court decides against member states. In this respect, the field's neglect of over-constitutionalisation, as emphasised by Grimm (2016), is relevant. In interpreting the Treaties, the Court may lock in certain policies, while precluding other policy options at the domestic and the European levels. Legal scholars are aware of the problem of overly constraining precedent, which only allows future deviations at the price of legal incoherence. Constitutional courts, Mark Dawson argues, 'not only rule … on the compatibility of legislation with constitutional rules but must take responsibility for imagining future alternatives. This is certainly a responsibility which the European Courts do not seem to consider themselves bound' (Dawson 2013: 20). The lock-in effect is even more severe for political decision-makers, since they cannot overrule constitutionalised case law. For example, when discussing the Services Directive in the mid-2000s, national treatment was not an option the European Parliament could pursue, as past case law on the freedom to provide services precluded such a restrictive interpretation (Schmidt 2015). Thus, Court rulings may largely conform to governmental preferences at the time when they are adopted -but they also constrain subsequent governments with potentially divergent preferences and, due to its constitutional status, case law can hardly be changed politically.
(iii) Finally, we need to consider the erga omnes relevance of Court jurisprudence. Courts decide disputes and the studies discussed above treat individual disputes as their unit of analysis. Assessing winners and losers for individual cases implicitly brings in the assumption that case law is only relevant inter partes, as is common in international law. If this was the case in the EU as well, we could clearly pinpoint the odds of member states having to live with the Court being more integrationist. According to Larsson and Naurin (2016: 395), in 57% of the cases, member states support the Court on the 'more Europe dimension' . The relevance of an ECJ ruling, however, is not restricted to the dispute at hand, since it has erga omnes effect. Once the Court clarifies a legal matter, this ruling is valid throughout the EU, and takes direct effect for national administrations, for the EU and member states' legislatures, as well as for member state courts. In fact, the significant work of Lindholm (2014, 2015) and  using network analysis shows the importance of precedent for the legal argumentation of the Court. As the interpretation of the Court is intended to settle the meaning of European law once and for all, it becomes more difficult to assess what it implies that in 43% of the cases it is at least not clear whether member states support the Court.
In sum, focusing on the interaction between member state governments and the Court only gives a partial answer about the importance of the Court. Given the erga omnes effect of rulings, we have to look beyond individual disputes to analyse how case law constrains member states. In this context, it is particularly relevant that over-constitutionalisation implies that in many disputes concerning policy the Court refers to the Treaty next to secondary law. As is the case for national constitutional courts, interpretations of the constitution become part of the constitution. Accordingly, if the ECJ interprets the four freedoms, EU citizenship, or European competition law in a certain way, the political institutions of the EU and its member states cannot deviate from this interpretation of supreme EU law. We now turn to the question of how case law shapes policy-making.

False expectations vs real impact
From a Political Science perspective, the question about how courts decide is particularly relevant because of their impact on policy-making. Ultimately, the yardstick against which to measure the importance of the ECJ in the process of integration is whether 'its considerations and doctrines become incorporated in the policy-making process' and whether 'the Court can promote distinct European policies and eventually shape legislation outcomes' (Wasserfallen 2010(Wasserfallen : 1129. Surprisingly, Political Science research on the (legislative) impact of the Court is scarce. It is here that Dorte Martinsen's (2015a) book breaks new ground. In the following, we focus on the Court's influence on (i) EU legislation and (ii) domestic legislation, and outline suggestions for further research.
(i) At the European level, the influence of Court jurisprudence on EU legislation is largely overlooked. Individual policy studies have repeatedly shown a significant impact of the Court's case law, in particular when used instrumentally by the Commission, e.g. to establish mutual recognition (Alter and Meunier-Aitsahalia 1994), to push liberalisation (Schmidt 2000), and to regulate defence procurement (Blauberger and Weiss 2013). In the general field of EU legislative studies, however, researchers tend to ignore the implications of case law. An example is the seminal book by Robert Thomson (2011) on the Council of Ministers. Being the result of one of the largest empirical exercises in EU studies, analysing the preferences of member states in EU decision-making, the index does not even contain an entry for the Court. Though over-constitutionalisation implies that the Commission's proposals for secondary law often consist of codification of case law, the field of EU legislative studies ignores this.
In her recent book, Dorte Martinsen (2015a, for an overview of major findings, see also Martinsen 2015b) tackles this huge gap and provides the most comprehensive and systematic study on the ECJ's legislative impact to date. In a nutshell, Martinsen develops a taxonomy of legislative responses to the Court's case law and analyses systematically the conditions of judicial influence on more than five decades of EU social policy-making. She concludes by rejecting any absolute claims about either a powerless or an 'ever more powerful Court': 'Judicial influence is not a question of whether but a question of degree' (Martinsen 2015a: 236). This conclusion is supported by the empirical finding that member states hardly ever override the Court's case law; they codify the case law mostly regarding technical issues; and they try to restrict the Court's impact in more contested areas through 'modification' (Martinsen 2015a: 95-7).
In its systematic approach, Martinsen's book sets a new standard and provides a blueprint for studying the ECJ's legislative impact beyond social policy. Nevertheless, we have two important caveats. First, the taxonomy of EU legislative responses to ECJ case law would be even more convincing without the category of 'modification' , since the latter ignores the fragmented nature of case law and blurs the distinction between override and codification. Case law is always about highly specific constellations and, therefore, necessarily fragmented rather than prescribing a full-fledged policy. Turning case-specific jurisprudence into general policy, then, does not amount to modification, but codification. Filling gaps left by and generalising from individual rulings is precisely what codification is about. This process always involves political signals as to whether existing case law should be interpreted more or less extensively in the future. And codification sometimes has a pre-emptive character when member states try to avoid further case law and regard political agreement on EU legislation as the 'lesser evil' (Schmidt 2000). In our view, such instances in which case law essentially pushes member states towards legislative agreement, however, are particularly strong examples of ECJ legislative influence rather than of 'weak' modification (Martinsen 2015a: 36). By contrast, if legislation truly contradicts the Court's jurisprudence, this is not modification either, but should be correctly categorised as override. Gareth Davies has made a similar argument and provided a clear criterion to distinguish override and codification, namely 'whether the legislation says things which the Court, if asked, would disagree with' (Davies 2016: 852).
Secondly, the over-constitutionalisation of EU law explains EU member states' limited but differentiated ability to adopt legislative corrections. Overriding the Court's jurisprudence via secondary legislation is impossible, if the former is based on an interpretation of the EU's Treaties. In these cases, the ECJ's 'pro-integrative rulings are effectively insulated from member state override' (Stone Sweet and Brunell 2012: 205); 'the idea of "legislative override" is a myth' (Davies 2016: 846). This is not to say that member states cannot signal their preferences to the Court -ex ante through written observations and ex post in the process of codification. Ultimately, however, it is up to the Court to be responsive or not: 'the Court's alignment to the other institutions, where it occurs, does so on an essentially voluntary/discretionary basis' (Hatzopoulos 2013). By contrast, the Court's interpretation of EU secondary law can be undone or amended via corrective legislation (Davies 2016: 850).
And finally, EU member states may legislate pre-emptively, before an issue becomes the subject of Court jurisprudence at all. Martinsen includes in her book the Working Time Directive where codification concerned case law on secondary legislation. By treating it largely on a par with the Patient Mobility Directive and the Posted Workers Directive, she fails to acknowledge the importance of over-constitutionalisation and to capture member states' limited but differentiated ability to correct ECJ jurisprudence through secondary legislation. In the specific case of the Working Time Directive, member states would have been able to override the Court, but they failed to reach legislative agreement with the European Parliament. Moreover, the existing directive already allowed generous possibilities of opt-out, including opting out of the constraints of case law. By contrast, where ECJ case law is constitutionally protected, secondary legislation can never overrule primary legal interpretation: legal hierarchy does not allow such 'modification' and overrule would require a unanimous Treaty change. Against this background of constitutionalised policy goals, future studies on the Court's legislative impact should compare more systematically cases in which the EU legislature has to deal with jurisprudence based on primary and on secondary law.
(ii) At the member state level, the Court's influence on domestic legislation has long been neglected or underestimated as well. Most Europeanisation and compliance studies focus exclusively on the transposition and implementation of European secondary legislation (Mastenbroek 2005(Mastenbroek : 1004. In this context, litigation in front of the ECJ or national courts has been analysed as a tool for ensuring member state compliance with EU legislation (Panke 2007). By contrast, Europeanisation studies largely neglected the ECJ's case law interpreting the Treaties. The prevailing view in the literature was that Court-driven 'negative integration' triggered weaker adjustment pressures for member states than positively prescribed EU policies (Knill and Lehmkuhl 2002: 258). Only recently has the interest in 'Europeanisation through law' grown. According to Treib (2014: 13), 'overcoming the focus on positive integration' is one of the most innovative features of current EU implementation studies.
Recent studies on Europeanisation through law largely confirm Conant's (2002) thesis of 'contained compliance' (Treib 2014: 13), but they also go beyond this finding in at least two respects. First, mere neglect is often unsustainable or even counter-productive for limiting the domestic impact of the Court's case law. Rather, member state legislatures may have to engage in systematic reforms precisely to shield domestic policies from further judicial intervention (Blauberger 2012). The underlying political rationale in these instances, then, is that 'it is better to define than to be defined' (Martinsen 2005(Martinsen : 1049. National administrations are particularly averse to the legal uncertainty of ECJ case law (Schmidt 2008). When the costs of legal uncertainty get too high, therefore, 'pro-activism' (Martinsen 2005(Martinsen : 1049 or 'anticipatory obedience' (Blauberger 2014) become the more likely political outcomes. Or, in the words of twentieth century Italian literature: 'If we want things to stay as they are, things will have to change' (Tomasi di Lampedusa 1960: 40).
Secondly, ECJ jurisprudence has another impact on domestic policy-making that is difficult to identify, yet highly significant: domestic 'non-decisions' (Bachrach and Baratz 1963). The Court's interpretation of EU fundamental freedoms constrains the set of domestic policy options that are still considered compatible with EU law. For example, the Court constantly holds that 'though in the areas in which the Community does not have competence, the Member States remain, in principle, free to lay down the conditions …, they must nevertheless exercise that competence consistently with Community law' (Case C-341/05 Laval, No. 85). As a consequence, even 'core areas' (Genschel and Jachtenfuchs 2014) of national legislative competence are not immune to the constraining effect of the Court's case law. Attempts to guard subsidiarity often founder. For example, in the context of the Brexit referendum, the constraints that EU law puts on domestic policy were widely discussed, as policy attempts to restrict EU migrants' access to benefits in the UK were heavily constrained under ECJ case law (Blauberger and Schmidt forthcoming). This stands in stark contrast to the large body of Europeanisation studies assuming that only secondary legislation directly impacts national policy-making. Yet compared to studying the transposition of EU directives, identifying and tracing this kind of domestic 'non-decisions' back to ECJ jurisprudence is particularly challenging and, hence, gets easily neglected (Töller 2010: 429f.). Moreover, what makes comparison difficult and obscures general Europeanisation patterns is the diversity of domestic policy implications stemming from negative integration. Compared to specific harmonised EU policies, general Treaty principles such as non-restriction or non-discrimination potentially collide with a much greater plurality of national policies, i.e. they affect very diverse member state policies through possibly similar mechanisms.
In sum, the ECJ's case law cannot dictate European or national policies, but its impact deserves greater attention. If we do not start with unrealistic expectations of a quasi-dictatorial judicial power, we can detect highly significant, yet subtle forms of judicial influence on policy-making in the EU's multi-level system. Rather than prescribing full-scale policies, ECJ case law involves considerable legal uncertainty and, thereby, often provides an incentive for policy-makers to take legislative action in the first place. Even if codification involves a good deal of gap-filling and translation of individual judgments into general rules, constitutionalised Court jurisprudence sets the frame on which the Commission bases its proposals and may create the very need for secondary legislation. And even if national policy-makers typically try to limit the impact of unwelcome ECJ rulings, 'containing' judicial influence may require systematic pro-activism. At the same time, domestic inaction does not rule out judicial influence either -rather non-decisions may constitute the most important, albeit difficult to detect, effect of Courtdriven negative integration.

Conclusion
Recent studies have significantly advanced our understanding of the ECJ. Yet when read in the context of the dichotomy between intergovernmentalism and supranationalism these studies easily lead us astray, as it appears that the Court's influence on European integration has been overrated. At the same time, much research approaching the EU from a comparative politics perspective fails to acknowledge unique features of the EU legal system and the significant policy impact of the Court.
It is the contribution of Grimm's book (2016) to coin the term overconstitutionalisation that captures a neglected feature of the Court's jurisprudence and its impact on European integration more generally. Studies of the Court's alignment with member state preferences in individual rulings underestimate the importance of precedent on subsequent rulings of European or domestic courts and, hence, the expansive effect of ECJ jurisprudence over time. And the implications of direct effect and supremacy of EU law have been largely ignored in EU legislative studies, where existing case law often forms the basis of the Commission's proposals. The over-constitutionalisation of EU Treaty law with its many policy goals provides an opportunity structure for private actors (Cichowski 2007;Kelemen 2011), the Commission (Blauberger and Weiss 2013;Schmidt 2000) and lower courts (Alter 2001;Stone Sweet and Stranz 2012) to pursue those policy aims -but it severely constrains member state governments to deviate from past policy choices and their interpretation by the ECJ.

Disclosure statement
No potential conflict of interest was reported by the authors.

Funding
The work of the TransJudFare project was supported by the Norface Welfare State Futures (WSF) Joint Research Programme [grant number 462-14-040].

Notes on contributors
Michael Blauberger is Associate Professor at the University of Salzburg and member of the Salzburg Centre of European Union Studies (SCEUS). He has published widely on European integration and Europeanisation through case law. His current projects focus on alleged 'welfare migration' in the EU and on democratic backsliding in EU member states. [michael.blauberger@sbg.ac.at]