2.1 Introduction

In 1998, Mattli and Slaughter declared that ‘Political scientists have discovered the European Court of Justice’ (Mattli and Slaughter Reference Mattli and Slaughter1998: 177). The assertion came after the first political scientists had entered the ongoing debate among legal scholars about the role of law and courts as drivers of European integration. The same year, Armstrong responded more critically to the newcomers: ‘But has it [political science] discovered law?’ (Armstrong Reference Armstrong1998). Twenty-seven years later, that annoying (to political scientists) question still hangs in the air.

This chapter argues that Armstrong’s provocation deserves to be taken seriously and that now is the time to do it. It first sets out to specify the question in a way that makes it useful: What is it about law that political science research on the Court of Justice of the EU (CJEU) should discover? It then goes on to investigate what approaches to law that the last thirty years of political science research on the CJEU has taken, and what remains to be discovered. While our focus is law in political science research on judicial politics, we demonstrate how our findings benefit also the broader interdisciplinary field of empirical legal studies (ELS) in EU law.

We argue that political science research on judicial politics in the EU should discover law to the extent that it is necessary and useful to achieve the following: (1) Develop and test theories of how the interplay between the Court of Justice, the EU’s political institutions, and relevant actors in Member States shapes EU law, (2) that are based on clear micro-level assumptions about drivers of and constraints on actors’ behaviour, and (3) that are grounded in realistic understandings of judicial practices (i.e., what judges actually do).

For the purpose of this chapter, we adopt a working model of the law corresponding to what may be described as the ‘lawyer’s view’. Understanding the practice of judicial behaviour and judicial decision-making requires us to consider law as lawyers understand it. Central to this view is the existence of a set of norms, rules, and principles, that is distinct from other social norms and whose legal character can be determined based on their source and the procedure by which they have been established (see, e.g., Hart 1994). According to the ‘lawyer’s view’, these norms, rules, and principles, along with the legal concepts used to construct, organise, and use them (Frändberg Reference Frändberg, Hage and Von Der Pforden2009), exist – law as fact – and form a discrete and independent system (Cotterell Reference Cotterrell1992: 38–43). This also includes legal institutions and the practices of the legal community, for example accepted methods of legal interpretation and argumentation.

The goal of political science research is not to determine ‘what the law is’ or the appropriateness of legal argumentation. However, the study of judicial law-making, how courts change the law, rests on and (at least implicitly) makes claims about what the law was before and after some judicial intervention. Similarly, law’s constraining effect on judicial behaviour hinges largely on how clear it is for the actors involved what the law ‘is’, and which arguments are perceived to be legal, acceptable, and persuasive.

Our chapter proceeds as follows. We first review the literature of the last three decades based on our criteria introduced above. We focus mainly on research relevant for understanding judicial law-making – the creation, modification, and application of rules and principles through selection and interpretation of legal sources and judicial reasoning – by the CJEU (or the Court). Our review broadly follows the development of the field chronologically. It starts with a brief look back at the discovery phase and the initial debates in the 1990s. This period was characterised by an impressive surge of creative theorising, not always matched by credible empirical research designs. Law was conceptualised as a mask and a shield (Burley and Mattil Reference Burley and Mattli1993), but most importantly as a constraint on judicial discretion (Garrett, Kelemen, and Schultz Reference Garrett, Kelemen and Schulz1998). It goes on to the debates following Carrubba, Gabel, and Hankla (Reference Carrubba, Gabel and Hankla2008) and the subsequent (re)convergence around a state of the art where the CJEU is seen as situated in a separation of powers relationship with the EU Member States, acting with bounded discretion under uncertainty (Larsson and Naurin Reference Larsson and Naurin2016) but with significant implications for EU and national level legislation and policy-making (Martinsen Reference Martinsen2015; Schmidt Reference Schmidt2018).

We then turn to more recent research that demonstrates how political science in different ways has become more nuanced in its approach to law and the decision-making of the CJEU. A stream of new studies has focused on carefully theorising and measuring significant legal concepts (e.g., legal complexity, consistency, and uncertainty), and on identifying and explaining legal innovations and doctrinal developments (e.g., deference, effective judicial protection). Recent research has also opened the black box of the Court itself, generating a range of insightful studies on judicial practices and the internal affairs of the Court.

Finally, we will argue that there is untapped potential in taking a micro perspective on judicial law-making and exploring the multidimensionality of legal cases. By breaking down the process of judicial reasoning into a sequence of choices in the hands of judges, political scientists, and ELS, researchers may find that there is not only law but also politics to be discovered in the details.

2.2 The Discovery: Law as Mask, Shield, and Constraint

Political scientists discovered the CJEU in the early 1990s when they saw that it had a role to play in the debate between the two dominating integration theories, neofunctionalism and intergovernmentalism (Burely and Mattli Reference Burley and Mattli1993; Garrett Reference Garrett1992). According to Mattli and Slaughter, that debate had in fact already been settled when they made their assertion about the discovery in 1998. Scholarship had converged towards a middle ground, where the discretion of the Court to act ‘within a sphere of law insulated from direct political interference’ was recognised along with a realisation that both legal and political factors imposed ‘significant constraints’ on judicial behaviour (Mattli and Slaughter Reference Mattli and Slaughter1998: 178). That same year, Garrett, Kelemen, and Schultz (Reference Garrett, Kelemen and Schulz1998) endorsed their proposition, developing a model where both legal precedent and external political preferences worked as constraints on judicial behaviour in Luxembourg.

The discovery phase also included important works by Alter (Reference Alter1996, 2001), Pollack (Reference Pollack1997, Reference Pollack2003), Stone Sweet (Reference Stone Sweet2004), and Stone Sweet and Brunell (Reference Stone Sweet and Brunell1998). In hindsight, the theoretical development that occurred during a few years in the mid-1990s was impressive. The discovery of the Court seems to have been a catalyst of scholarly creativity and engagement. It is notable that the EU literature at the time was developed largely by American scholars and in tandem with the US judicial behaviour literature. As a result, how European judicial politics scholarship has approached and measured law is heavily influenced by American political science and legal thinking about law.

The theorising in these early studies had the elements of high quality political science research. It demonstrated clarity with regards to the micro-level preferences and constraints of key actors: judges are assumed to have interests pertaining to both institutional status and power as well as substantive policy preferences. But they are also ‘socialized to understand themselves as agents and servants of the law’ (Mattli and Slaughter Reference Mattli and Slaughter1998: 196). The Court is constrained, on the one hand, by the demands of legal reasoning, ‘the weight of text and precedent; the elemental requirements of precision, clarity, and determinacy in rule interpretation and application’ (Mattli and Slaughter Reference Mattli and Slaughter1998: 198). On the other hand, it was also perceived as constrained by the necessity of safeguarding its social legitimacy and to anticipate and avoid hostile political reactions.

Notably, Mattli and Slaughter politely conceded to the legal community that ‘the language of preferences and constraints is harshly instrumental, likely to jar and even offend the sensibilities of many judges (and lawyers). Judges are likely to describe their own thought processes in reaching decisions based on deeply internalized notions of both the idea and the ideal of the rule of law’ (Mattli and Slaughter Reference Mattli and Slaughter1998: 197). However, other than the nod towards the world of legal and judicial practice, there was little attempt in these studies to model or measure the content of the judgments or the decision process of the judges. The court remained a black box with a single mind: more Europe (Pollack Reference Pollack2013).

The empirical research designs were also relatively limited compared to today’s standards. While some of the work demonstrated the qualities of case study-based research, including a deep understanding of important events and cases, it also came with well-known limitations regarding external validity and theory testing. For example, landmark cases, where the Court makes important advances in case law, are central to this research, but they are rarely systematically compared to cases where the Court could have made a brave new move but chose not to.

Where was law in this? Several distinctive conceptualisations of law were present in the political science literature in this period. Burley and Mattli (Reference Burley and Mattli1993) famously developed the notion of law as a mask and shield. While law as a shield emphasises the ‘independent “non-political” rationale’ of judicial decision-making, ‘which must be met by a counterargument on its own terms’ (Burley and Mattli Reference Burley and Mattli1993: 72), law as a mask builds on neofunctionalist ideas about the role of technical expertise for European integration, focusing on the obscure nature of legal discourse (a frustration also for many political scientists). The Court of Justice’s development of EU case law had been able to fly under the radar of the Member States, according to this argument, partly because governments simply did not understand what the judges were saying (for a similar argument on obfuscation at the US Supreme Court, see Owens et al. Reference Owens, Wedeking and Wohlfarth2013). Latter research has argued that law as a mask may be a thing of the past, as politicisation of the EU has increased and as few close observers of the EU today underestimate the political potential of the Court’s decisions (Blauberger and Martinsen Reference Blauberger and Martinsen2020; Larsson et al. Reference Larsson, Naurin, Derlen and Lindholm2017: 881; but see Ovádek Reference Ovádek2022). The law-as-mask analogy introduced an important observation about law, namely that legal and political discourse are different in significant regards. However, while Burley and Mattli (Reference Burley and Mattli1993: 70) focused on how the Court can potentially use the ‘inexorable logic of law … to camouflage controversial political decisions in “technical” legal garb’, they paid less attention to when that legal logic and garb is warranted from a legal perspective.

Law as a constraint was subsequently most clearly developed by Garrett, Kelemen, and Schultz (Reference Garrett, Kelemen and Schulz1998), and remained the most important conception of law in the political science research on the CJEU in the following decade. In this notion, law is an external limitation on judicial decision-making contributing to defining the discretion of the court vis-a-vis the Member States. The Court is expected to provide answers to legal questions that are consistent with its existing jurisprudence. As Garrett, Kelemen, and Schultz (Reference Garrett, Kelemen and Schulz1998: 157) point out, if the CJEU’s ‘jurisprudence were to change frequently from case to case in response to pressure from the actors involved … the court would surely lose legitimacy’. The authors claim that the Court is more likely to reach decisions that have adverse effects for Member States if existing precedent and legal norms supporting such a decision are ‘clear’. While highlighting that precedent’s clarity factors into the Court’s decision-making, Garrett, Kelemen, and Schultz’s theoretical model omits that the Court itself to some extent determines such clarity. The CJEU may consciously avoid the creation of clear precedent to pre-empt future scenarios in which the Court is pressured into decisions that provoke conflict with Member States. Hence, law may act as a constraint, but it is the Court itself along with the EU legislator that tightens the bolts.

2.3 The Second Debate: Estimating the Boundaries of Judicial Dscretion

Despite Mattli and Slaughter’s (Reference Mattli and Slaughter1998: 178) attempt to close the book on the neofunctionalist–intergovernmentalist debate already in 1998, which they believed had ‘reached the limits of its usefulness’, the debate was far from over. It took a new and unusually heated turn after the publication of Carrubba, Gabel, and Hankla (Reference Carrubba, Gabel and Hankla2008, see also Carrubba, Gabel, and Hankla Reference Carruba, Matthew and Hankla2012; Stone Sweet and Brunell Reference Stone Sweet and Brunell2012). For a while, law had very little to do with it.

At the core of the debate was the question of how much discretion the judges in Luxembourg enjoyed from the Member States and what factors determined the limits to that discretion (threat of override and non-compliance). The debate resembles previous debates within the American judicial politics literature. The neofunctionalist position (similar to the attitudinalists such as Segal Reference Segal1997: 31) depicted a court with few limits, whether from law or politics. According to Stone Sweet and Brunell, the CJEU has no reason to take Member State governments’ preferences into account other than at the margins, because the Court is ‘effectively insulated’ from override, and non-compliance is not a problem the judges care much about (Stone Sweet and Brunell Reference Stone Sweet and Brunell2012: 205). As a result, the CJEU had been able to complete its ‘judicial construction of Europe’ largely undisturbed by external constraints, with a zone of discretion that was ‘close to unlimited’ (Stone Sweet Reference Stone Sweet2004: 9).

Carrubba and Gabel, on the other hand, argued that Member States grant the Court just enough discretion to allows them to create the credible commitments they need to co-operate efficiently (Carrubba and Gabel Reference Carrubba and Gabel2015; Carrubba, Gabel, and Hankla Reference Carrubba, Gabel and Hankla2008). How the judges in Luxembourg engage with law and legal reasoning has little to do with the outcome of importance, which is whether Member States comply with their commitments. Normative aspects of law as a shield, or judges’ professional expertise in legal reasoning, are not necessary to explain compliance, which is driven by the threat of sanctions by other Member States. The cases brough by litigants to the Court work as a fire alarm alerting the Member States that something fishy is going on. The Court’s only role is to provide a venue for the Member States to clear out their differences (‘information clearinghouse’).Footnote ¹

Carrubba and Gabel provide a very slim role to law and judicial decision-making. They do so not because they are unaware that judges also engage with law, involving legal sources, case facts, and legal reasoning. The power of Carrubba and Gabel’s argument comes from the fact that they do not need to theorise what the court is doing with the law to demonstrate the logics of their theory of how international courts like the CJEU provide important functions that makes international co-operation possible. However, a cost of simplification in this case is that the model is not grounded in a realistic understanding of what judges do. Consequently, the judicial proceeding becomes a theatre play, where the real questions asked are not those that are voiced in the hearings and put on paper in the opinions and judgments. While the Court’s reasoning seems to focus on whether a Member State action is legal or not, what the Court is really looking for is information about whether the other Member States find the action tolerable from a material policy perspective.Footnote ² While Carrubba and Gabel’s work has been groundbreaking in the international relations and comparative judicial politics literature, both theoretically and empirically, the detachment from judicial practices makes it a hard sell to ELS.

In Carrubba and Gabel’s research design, too, law entered the analysis merely as a control variable in the form of the Advocate General’s opinion, which was presumed to be a proxy for the legal merits of the case. Unfazed by the political mood in Member States, the Advocate General (AG) would provide answers to legal questions that are consistent with the CJEU’s existing jurisprudence, mitigate the legal uncertainty that had sparked the question in the first place, and consider the impact of their answer on the future development of EU law. Hence, once the researcher knows what the Advocate General proposed in their opinion, the judges’ legal considerations are accounted for in the analysis and any deviation in the CJEU’s ruling from the Advocate General’s opinion must stem from the political pressures felt by the Court (see also similar work by Larsson and Naurin Reference Larsson and Naurin2016).

Notably, not all political science research studying the impact of the CJEU’s judgments on Member State sovereignty or reactions from the legislative institutions of the EU showed the same disinterest in the finer details of law. Martinsen’s (Reference Martinsen2015) study of the CJEU’s influence over EU social policy uncovered ‘what becomes of law after it leaves the hands of the judiciary and enters legislative politics with its battles, dialogues, and compromises’ (Martinsen Reference Martinsen2015: 3). Her work extended political scientists’ conceptual toolkit, finding nuance beyond the dichotomous distinction between compliance and non-compliance. Negotiating in the EU’s legislative process over their response to the CJEU’s decisions, Member States may ‘codify’ judge-made law in legislation, ‘modify’ what the CJEU had intended to achieve, ‘override’ judicial interpretations of EU secondary legislation, or simply fail to respond at all, causing legal uncertainty (Martinsen Reference Martinsen2015: 35–6). The effects of the ‘codification’ of the case law created by the CJEU also features prominently in Schmidt’s (Reference Schmidt2018) work. Although the CJEU resolves legal questions that typically involve only one or two Member States, its answers set legal precedents that apply to all Member States. Emphasising the precedential nature of the CJEU’s case law and its erga omnes effects throughout the EU, Schmidt (Reference Schmidt2018: 47) argues that the CJEU does not need to regularly constrain national sovereignty to have a lasting impact on politics in the EU and Member States – ‘even infrequent activist rulings go a long way’.

Eventually, the literature converged on a broad agreement about a court with substantial discretion, with strong preferences for European integration, limited by its political and social environment to varying degrees. Its discretion was seen to depend on the salience and conflicts of interest generated by the cases, and on the institutional rules governing collective action among the Court’s interlocutors (e.g., Blauberger and Schmidt Reference Blauberger and Schmidt2017; Larsson and Naurin Reference Larsson and Naurin2016; Martinsen Reference Martinsen2015). How the judges reached their conclusions as outlined in their legal reasoning, however, was not an object of research. While the conception of law as a mask and constraint in earlier research had pointed to a feature of judicial reasoning with implications for the Court’s discretion, it ostensibly did not invite scholars who built on this work to disentangle the technical jargon or investigate whether the reasoning included significant choices for the judges.

The limited attention given to law and judicial practices in the literature likely contributed to the perception among legal researchers (even those interested in ELS) that political science research was out of touch with the real world of judicial decision-making and the inability of this research to answer legally relevant questions. Similar concerns about political science and law have featured in relation to the US judicial politics literature. It underlines the potency of Armstrong’s question.

However, at the same time, studies of citations of case law pointed to new opportunities for testing political science theories, by taking more seriously the content of the decisions of the CJEU and the legal reasoning of the judges (Frankenreiter Reference Frankenreiter2017; Larsson et al. Reference Larsson, Naurin, Derlen and Lindholm2017). These studies were inspired by work from other courts (Lupu and Fowler Reference Lupu and Fowler2013; Lupu and Voeten 2014) and by legal researchers (Derlén and Lindholm Reference Derlén and Lindholm2014; Šadl and Hink Reference Šadl and Hink2014). They demonstrated that significant political science questions could be addressed by looking more carefully at the reasoning of the Court.

Larsson et al. (Reference Larsson, Naurin, Derlen and Lindholm2017) used network analysis to calculate measures of the ‘embeddedness’ of judicial decisions in previous case law, finding that the Court put more effort into its reasoning when taking politically sensitive decisions that constrained the sovereignty of EU Member States. Still, the judicial practice of citing case law was set in a familiar frame of Court–Member State interactions. The purpose of embedding the decisions in case law was assumed to be to increase the zone of discretion of the Court by fortifying law as a constraint when it aligned with the judges’ preferences for More Europe.

Frankenreiter (Reference Frankenreiter2017) connected citations to case law to the political ideology of the opinion writers, finding that judges appointed by Member State governments with a pro-European leaning were more likely to cite previous cases written by judges with a similar background. The study followed previous work that had started to open up not only for more theorising about individual judges’ heterogeneous policy preferences (Malecki Reference Malecki2012), but also for considering the link to Member States through the appointment procedure (Kelemen Reference Kelemen2012). While Frankenreiter engaged with one expression of legal reasoning (citations) he did so in order to reveal ideological preferences, placing law firmly in the background.

2.4 The Present: Political Science Research with a Better Sense of Law as Craft …

More recently, a stream of new studies shows that political science research is coming closer to the nuts and bolts of CJEU decision-making, without losing in stringency with regards to general theory and credible research designs.

Two recent studies have made important contributions into understanding the Court’s approach to the rule of law, one of the most important challenges for the EU in recent years triggered by democratic backsliding in some Member States. Ovádek (Reference Ovádek2022) studied both the background to and the implications for doctrinal development of the Court’s landmark judgment in Portuguese Judges (C‑64/16). The study uses political science theory to explain the judgment and its subsequent impact. While recognising that judicial law-making is constrained by the cases a court receives, Ovádek shows how the CJEU decided to make a lot more of Portuguese Judges than anyone expected. In a case that was not strategically driven by Euro-lawyers, that flew under the radar of the Member States (only one submitted observation), the Court reformulated the question asked by the national court, ignored the Commission’s suggestion that the case was inadmissible, and went much further on the merits than the Advocate General in their opinion in terms of expanding the Court’s jurisdiction into core matters of state institutional design.

Manriquez and Pavone (Reference Mandujano Manriquez and Pavone2024) develop a new measurement strategy for studying judicial innovations. They define a legal innovation as a proposal for establishing a new legal basis or principle for rule of law enforcement, and/or making such innovations ‘bite’ by sanctioning Member States’ actions as infringing such innovative legal constructs. By a combination of citation analysis, hand coding, and the use of regular expressions they identify when the Court innovates in its case law. They investigated all ninety-six rule of law cases decided by the Court after 2010, finding that sixteen of those included legal innovations. Furthermore, they also study whether the Court might have been inspired by earlier suggestions by the Commission in its observations and extra-legal writings. Manriquez and Pavone conclude that it is the Court itself that has been the main innovator in the rule of law area in recent years, thus taking on a crucial role in an area where the other EU institutions were struggling to take decisive action.

These two studies have in common serious engagement with legal doctrine and the judicial reasoning of the Court in a large number of cases, firm anchoring in general theory about European integration and agenda setting (Manriquez and Pavone Reference Manriquez and Pavone2024), and legal mobilisation and separation of powers models of judicial politics (Ovádek Reference Ovádek2022). They are both built on credible research designs, including a systematic selection of cases, rather than looking only at the most visible landmark cases, and a mixed methods empirical strategy.

Two other recent studies based on IUROPA data show how political science work can include both an understanding of the decision-making process of the CJEU and a sense of the legal purpose of a court of precedent like the CJEU, in contrast to the more functionalist theories of international courts as credible commitment devices found in the International Relations (IR) literature. Lindholm, Naurin, and Schroeder (Reference Lindholm, Naurin and Schroeder2025) argue that negative references to Member States observations – instances where the CJEU explicitly signals disagreement with the legal arguments by Member States – provides the Court with a tool for increasing the precision of its case law. Their empirical analysis indicates that negative references are used both to boost its legitimacy and to specify not only what the law is but also what it is not.

Schroeder (Reference Schroeder and Lindholm2023) demonstrates empirically that the CJEU is more likely to leave it to national courts to decide on legal questions when these are narrower and more contextual, affecting only a single or a smaller set of Member States. One important aspect of the theory is the uncertainty about the future legal and policy implications of the Court’s decisions and the information the Court can collect from other actors to mitigate that uncertainty. Analysing all preliminary reference cases lodged with the Court between 1995 and 2011, Schroeder shows that the Court is more likely to clarify legal rules and principles in a more precise manner when it senses a stronger need to safeguard the uniform application of EU law. This work builds on and contributes to previous political science literature, but also to the work by legal scholars on the deference doctrine (Zglinski Reference Zglinski2020).

2.5 … and Judicial Practice

For many years, the inner workings of the CJEU was a black box to political science. In his review of the literature in 2013, Pollack argued that ‘a virtual consensus exists among otherwise diverse disciplines and otherwise hostile schools of thought that the Court should be studied as a unitary actor with a consistent, decades-long preference for European integration’ (Pollack Reference Pollack2013: 1265). Opening that box gives political scientists access to the production of substantive law. Some of the procedural elements may also inform researchers about important legal aspects of court judgments, such as whether new important legal issues are raised in a case. Furthermore, acknowledging the significance of the practice and procedures of judicial decision-making helps political scientists communicate research findings with judges and legal researchers. Understanding the judicial practices also helps with generating new research questions, developing new theories of relevance to political science and ELS, and coming up with new ideas for empirical operationalisations to test theories.

Malecki (Reference Malecki2012) and Kelemen (Reference Kelemen2012) are early examples of research looking into the delegation of decision-making within the CJEU to smaller panels of judges. Fjelstul has broken new ground in this area by theorising the role of the chamber system and analysing the consistency and productivity of the Court (Fjelstul Reference Fjelstul, Gabel and Carrubba2023; Fjelstul, Gabel, and Carrubba Reference Fjelstul, Gabel and Carrubba2023). This research shows that the increasing use over time of the practice of delegating decision-making to smaller panels has fulfilled its purpose in terms of productivity, reducing case duration and the Court’s backlog.

Moreover, highlighting the link between judicial procedures and the substance of law, this research also indicates that increased productivity may come with serious costs in terms of inconsistency when cases are decided in smaller chambers. Fjelstul uses a case space model and machine learning techniques to generate predictions about the impact of the size of the panel on the consistency of outcomes in infringement cases. Inconsistency, in this perspective, occurs when a party that won the case in one chamber could have lost in a differently composed chamber. The findings contain worrying signs of inconsistency in the Court’s application of the infringement procedure, in particular under legal uncertainty (when the case facts are not clearly favouring one party) and preference heterogeneity among judges (Fjelstul Reference Fjelstul, Gabel and Carrubba2023). In the empirical analyses, Fjelstul also designs new ways to measure legal complexity useful for future work (see, e.g., Schroeder Reference Schroeder2025).

Other studies look more closely at the role and functioning of key members of the Court, including the Reporting Judge (Boulaziz Reference Boulaziz2025; Cheruvu Reference Cheruvu2019; Hermansen Reference Hermansen2020), the President of the Court (Hermansen Reference Hermansen2020) and the Advocate General (Carrubba and Gabel Reference Carrubba and Gabel2015; Frankenreiter Reference Frankenreiter2018; Makszimov Reference Makszimov2024). While the CJEU is non-transparent in terms of judicial votes and does not allow separate opinions, it publishes information on who is the Reporting Judge. The Reporting Judge is case manager with important responsibilities from the early stages of the process to the final writing of the judgment. They make proposals to the General Meeting of the Court concerning key procedural questions (chamber size, whether more background research is needed, whether an AG opinion or an oral hearing is needed), they draft the first outline and manage revisions of the judgment during the deliberations in the chamber (regardless of whether they are in the majority or not, see Dunnoff and Pollack Reference Dunoff and Pollack2017).

Hermansen (Reference Hermansen2020) studies the assignment of Reporting Judges by the President of the Court. The assignment of the Reporting Judge is potentially a powerful tool in the hands of the President for influencing the direction and coherence of the Court’s case law (and possibly for building an internal coalition favourable to the President’s re-election). Looking at almost 10,000 case allocations over a forty-five-year period, Hermansen finds that these assignments are far from random, but strategically tailored to satisfy the competing demands of legal consistency and external social-political legitimacy. The President is more likely to assign cases with substantial legal uncertainty to judges appointed by governments with a moderate ideological position (on the economic Left–Right scale). Hermansen also finds that case allocation to some extent contributes to the specialisation of judges in specific areas of law, despite the Court’s explicit denial of practicing such specialisation.

This research brings new light to one of the most important procedural decisions in the Court. It pictures a President who needs to balance the legal professional expectations on the Court in terms of developing consistent case law, while at the same time managing its political environment. It is cognisant of the fact that the judges on the Court are both legal professionals and appointees of the governments of the Member States, pointing to the importance of taking seriously the implications of the fact that the Member States appoint and re-appoint the members of the Court (see also Cheruvu Reference Cheruvu2024; Frankenreiter Reference Frankenreiter2017, Reference Frankenreiter2018; Hermansen and Naurin forthcoming; Kelemen Reference Kelemen2012).

Boulaziz (Reference Boulaziz2025) studies gender-based selection and performance biases at the Court. She finds that female judges are more often assigned what she calls ‘the judicial housework’ at the Court, while male judges are assigned to more prestigious groundbreaking cases. Moreover, she finds that female judges complete their cases faster than their male counterparts. Cheruvu (Reference Cheruvu2019) demonstrates that who gets assigned as the Reporting Judge in combination with the Court’s French-language mandate matters also for the efficiency of the Court. His study of more than 10,000 judgments finds that judges who are natively French speaking complete their cases as Reporting Judge about a month quicker than non-francophone judges.

These studies show that political science is getting closer to the factory floor where EU case law is being produced, complementing the work by legal researchers on the actors and processes of the internal decision-making of the Court (e.g., Kim Reference Kim2022; Krenn Reference Krenn2022; Šadl and Sankari Reference Šadl and Sankari2017; Wijtvliet and Dyevre Reference Wijtvliet and Dyevre2021). Opening the black box has been facilitated by the IUROPA CJEU Database, which includes complete and research-ready information on a range of procedural events and decisions as well as positions and background information of key actors involved in the Court’s decision-making (Brekke et al. Reference Brekke, Fjelstul, Hermansen and Naurin2023).

2.6 The Future: The Politics Is In the Details

A constructive response to concerns about a perceived detachment of political science research from law and judicial practice is the case space approach to judicial decision-making. It includes an explicit recognition that judicial law-making is different from policy-making in the executive or legislative branches in that its incremental and constrained by the facts of the case (see Clark Reference Clark2019; Kornhauser Reference Kornhauser1992a, Reference Kornhauser1992b; Lax Reference Lax2007 on theorising the decision-making of courts of precedent in the United States). Political scientists interested in explaining the decision-making of the CJEU should develop theoretical frameworks that capture how judges make law one case at a time. Each case reaching the CJEU may be a vehicle for law-making – for defining rules and principles and developing doctrine – but these cases come with specific sets of factual circumstances that frame what is possible.

Recognising that judicial law-making is distinct from policy-making also invites the political scientist to consider the multidimensionality of legal problems, rather than sticking to simple policy scales such as Left–Right or More–Less Europe. Judges approach cases as ‘bundles of facts’, and adjudication requires determining which legal and factual dimensions are relevant to consider before a decision is possible. This goes for both the specification of rules, the balancing of principles, and eventually for determining whether a challenged action was on the right side of the law. While thus attentive to the micro-level details of adjudication, what is liberating from a political science perspective is that it is up to the researcher to theorise whether judges’ decisions (e.g., the selection of relevant factual and legal dimensions) are rooted in political ideology, judges’ concerns about societal impact or social legitimacy, or different views on legal methodology or philosophy (see Lax Reference Lax2007: 592).

Equally comforting to the political scientist is that this way of thinking about judicial decision-making invites theorising about strategic behaviour. The fact that adjudicating cases requires addressing multidimensional problems opens the door to classic political science theory about agenda setting, including ‘conflict expansion and contraction’ (Schattschneider Reference Schattschneider1960) and ‘heresthetics’ (Riker Reference Riker1986). Judges and lawyers may strategically emphasise those dimensions of the case that make it more likely that their preferred outcome prevails. ‘What is this case about?’ is a legal and a factual question that is open to manipulation. Which dimensions that become decisive for the case may determine the winning coalition within the court as well as the outcome in a broader sense, including both precedent and policy impact.

This approach provides opportunities for analysing how judges’ and other actors’ preferences and ideas are promoted through law. In Clark’s (Reference Clark2019) terminology, law is the material (clay) that judges use to reach their goals (pottery). Here is not the place to develop a detailed case space model of the preliminary reference procedure, but a brief example may illustrate the potential for political scientists to discover law in a way that is likely to be constructive for interdisciplinary synergy and generating new questions and research designs.

In his important work on free movement in EU law, Zglinski has demonstrated how the CJEU over time has shifted its attention from questions concerning the scope of application of rules to justification of restrictive measures (Zglinski Reference Zglinski2020). In free movement cases, the Court’s reasoning often comes in the following sequence: First, the Court must decide whether a restriction on free movement has occurred in the case at hand. If it finds that this is the case, it goes on to determine whether that restriction was justified or not (often by means of a proportionality review). In practice, the principle of free movement is then balanced against other public values that the Member States seek to defend by means of restrictive measures.

The sequential structure of the Court’s reasoning invites theorising of judicial heresthetics: We may assume that judges will anticipate what the implications are of closing or opening the door to the second justificatory stage. For judges that are at least partly motivated by nurturing the social legitimacy of the Court, and therefore sensitive to the policy implications of its judgments, that decision may have profound consequences. Allowing the case to go through to the justification stage may force/invite the Court to make decisions in politically highly sensitive matters.

A closer look at the scope of the application stage reveals that it includes several decisions that may be used to close or open that door. In case space model terminology, the Court defines the relevant factual dimensions, stipulates the rule by specifying thresholds, and places the case inside or outside of the rule. A strategic court that anticipates trouble – should it be forced to balance sensitive public values against free movement – may search for ways to stop the case moving to the justificatory stage.Footnote ³

Zglinski’s work shows that the Court’s room for manoeuvre in this regard has decreased over time in the area of free movement of goods. He found that defining the scope of free movement was the main occupation of the Court for decades, but that gradually more focus was directed towards the justification stage. The central problem for the Court ceased to be whether national laws fall within the scope of the free movement rules. Instead, it became ‘can these laws be justified and, most importantly, are they proportionate?’ (Zglinski Reference Zglinski2020: 14). One important reason for the shift in attention is that – due to the Court’s own case law – the scope of the rules has become so encompassing that they now cover most of the cases that arrive at the Court.

One implication of this shift is that the types of problems that the Court needs to deal with have also changed, often including politically sensitive rights and interests (Zglinski Reference Zglinski2020: 14). At the same time, there may be more opportunities for strategic judicial behaviour at the justificatory stage. In the preliminary reference procedure, the CJEU may choose not to decide on the disposition of the case, by deferring to national legislative authorities or national courts. In fact, there are several options for the Court if it allows the case to go through to the justification stage. The CJEU may:

1. (a) Go all in, by fully specifying the relevant dimensions of the case space, drawing the line and placing the case (what Zglinski calls ‘no margin of appreciation’, e.g., Cassis de Dijon (C-120/78)). Although formally, it is for the national court to decide the case at hand, in practice the CJEU may go a long way towards doing that.

2. (b) Go some way in, by defining the relevant dimensions and drawing a line, without placing the case. The latter happens when the court determines what the case is about (factual dimensions) and identifies a threshold or substantive criteria for national courts to take into account when deciding ‘cases like these’, without determining on which side the specific case at hand falls (what Zglinski calls ‘partial margin of appreciation’, e.g., Food additives Denmark (C-3/00)).

3. (c) Go lightly in, by indicating the relevant dimensions, but leaving it to national authorities to draw the line and place the case (what Zglinski calls ‘full margin of appreciation’, e.g., Omega (C-36/02)).

The point here is that there is little to lose in terms of political science theory by coming closer to the everyday professional activities of adjudication and judicial reasoning. Specifying the micro-level choices that go into deciding cases is fully compatible with theorising about judicial decision-making as strategic choice, both in terms of agenda setting and disposition, and of case law as the product (add ‘partial’ if you like) of judges’ preferences (whatever you think they are). There is plenty of politics to be explored in these details.

However, an important remaining challenge is that gathering large-N data on the more fine-grained elements of law and judicial reasoning requires a lot more than looking for winners and losers in the operative part of judgments. This is likely one explanation why there are still relatively few large-scale empirical applications of the case space model (Clark Reference Clark2019 being a pioneer in this regard).Footnote ⁴ Nevertheless, knowing that these choices exist and are expressed in the reasoning, that is, in the text of the judgment, is a first step. Furthermore, recent work by Schroeder and Lindholm develops a useful machine learning approach to splitting judgments of the CJEU into more fine-grained units of analysis that they refer to as ‘issues’; ‘blocks of text within judgments that address distinct legal questions’ (Schroeder and Lindholm Reference Schroeder and Lindholm2023: 165).

2.7 Discussion

One could frame these suggestions for future research as political science with a sense of law and legal practice. More importantly is to acknowledge that such research is transcending the disciplinary boundaries and pointing towards a truly interdisciplinary ELS. In our view, it is largely a myth that legal researchers and political scientists are interested in different things. Most of the time, political scientists and legal researchers in ELS raise questions about the origin and impact of law and legal change that call for convergence with regards to conceptualisations and measurements strategies.

Is there a risk that political science research that focuses more on the details of judicial reasoning and practices becomes less attentive to significant macro questions about the separation of powers in the EU and the role of the CJEU for European integration? When you dive deeper, it is easy to get stuck in the nitty gritty and to forget to look up to see the bigger picture. This is where large-scale identification and quantification of legal concepts, methods, and practices come in (Kim and Schroeder 2024). Careful conceptualisation and measurement strategies allow for taking the reality of law and judicial practice seriously while developing and testing general theories about law and politics.