Or Brook¹ and Barry Rodger²

1 Introdution

The past twenty years has witnessed substantial growth in the systematic analysis of judgments issued by courts, especially in the North American context (Hall and Wright 2008; Weinshall and Epstein 2020). Such studies bring the rigor of social science to our understanding of case law and judicial behaviour, by generating “objective, falsifiable, and reproducible knowledge about what courts do and how and why they do it” (Hall and Wright 2008). Nevertheless, the empirical study of EU law in general, and of the operation of European courts in particular, is still greatly underdeveloped. The legal education and training of European law students, the metrics for evaluating success in European legal academia, the traditional European legal culture and the roles it assigns to legislators and the courts, and the interplay between EU and national lawmakers have all been suggested as possible sources for the limited embracement of empirical legal research in Europe.³

Since the 1990s, some ground-breaking empirical studies have sought to explore the operation of the EU Courts (the General Court (GC) and the European Court of Justice (ECJ)).⁴ Yet, the limited systematic investigation of the operation of national courts applying EU law is striking. Only very few studies have examined the operation of national courts and their impact on the EU model of judicial federalism, despite the fact that national courts have become the primary judicial forum in which EU law is applied (Wells 2016; Halberstam 2021).

In addition to the general obstacles for the systematic analysis of court judgments in Europe described above, the EU-wide and comparative-empirical study of national courts applying EU law and judicial federalism encounters unique challenges tied to the limited publication and availability of national courts’ judgments, restrictions on the systematic exploration of such data, language barriers, diverging legal traditions, and considerably different procedural and institutional settings. As elaborated below, the EU’s ‘vertical’ federalist character entails that the administration of EU law — including the central role of the judiciary — mostly relies on the resources and institutions of its Member States, which are free to devise their own procedural rules and institutions to undertake this task. Known as the principle of procedural and institutional autonomy, this means that although the Member States are bound to apply the EU substantive law provisions, the national institutions and procedures applying such laws may considerably diverge across the different Member States within the EU. The EU’s governance model stands in contrast to the United States’ ‘horizontal’ federalist framework, in which both federal and state governments retain the authority to legislate, execute, and adjudicate policies and laws independently. Consequently, US federal and state antitrust laws exhibit a greater degree of autonomy from one another, diverging not only in institutional and procedural aspects but also in substantive legal provisions.

The above challenges informed the construction of the Competition Law Judicial Review Database, the first comprehensive study⁵ mapping out the national judicial review of competition law (antitrust)⁶ public enforcement in the European Union (EU) (2004–2021). Following the collective efforts of twenty eight national teams of experts, the Database encodes case- and jurisdiction-level information from all publicly available national judgments reviewing the application of the EU prohibitions against anti-competitive agreements (Article 101 of the Treaty on the Functioning of the European Union (TFEU)) and abuse of a dominant position (Article 102 TFEU) and of the national equivalent provisions of those EU rules by national competition authorities (NCAs) between the entry into force of Regulation 1/2003 in May 2004 and the end of April 2021 (5707 judgments). The database, including an online coding book and analysis tools, is available at https://www.mappingcomplawreview.com/.

National courts play a special role in the enforcement of competition laws across the EU, as they constitute the main forum for the review of enforcement decisions taken by NCAs.⁷ Courts must secure the effectiveness of the competition authorities’ enforcement and decision-making, including the due protection of affected parties’ rights. This is not a simple task, especially as the application of the competition rules often calls for a complex economic and legal assessment.⁸ At the same time, beyond the trite assertion that judicial review seeks to protect the rule of law, European competition law and scholarship have offered scant guidance on the functions of judicial review. The effectiveness of judicial review within the EU legal order has often been taken for granted (as observed by Geradin and Petit 2012), and very little is known about national courts’ systems for review of decisions enforcing EU (and/or domestic competition law), and how the institutional and practical divergences in these systems across the EU might have an impact on the interpretation and enforcement of the same set of rules.

The creation of the database, as well as our experiences in building and managing this project shared in this article, aspire to fill two main gaps in existing empirical legal scholarship:

The first gap is EU competition law-centric, aiming to shed light on the context of national enforcement of EU competition law. Thus far, the scholarship on the judicial review of EU competition law enforcement has mostly focused on the operation of the EU Courts — the GC and CJEU — reviewing the enforcement actions taken centrally by the European Commission. The limited number of studies focusing on national courts have mostly followed ‘traditional’ legal-doctrinal methodologies, examining the rules (Temple Lang 2014; Kokkoris and Lianos 2019) and jurisprudence on national courts, for example, via the lenses of coherence (Sauter 2016) or the impact on the rule of law (Bernatt 2019). Very few empirical-systematic studies have explored the issues around national judicial review, and they have mostly been limited to consideration of a single or a very limited number of legal systems (Rodger et al 2024). As a result, although the basic institutional structures of the various national judicial systems involved in EU competition law may be transparent, little is known of their actual functioning across the EU, about any potential inconsistencies in the application of the EU law rules applied by them, and regarding the challenges they may pose for access to justice. As discussed in detail below, our project investigated the hypothesis that, in the absence of harmonizing provisions, the standards, scope, and intensity of judicial review are likely to vary significantly across EU Member States. This variability, in turn, is likely to lead to divergent review processes and outcomes, undermining the objectives and foundational principles of the decentralized enforcement system of EU competition law.⁹

The second gap this article addresses pertains to the challenges of collecting judgments and applying systematic content analysis of legal text (‘coding’) in the context of an empirical comparative law project, using cross-country legal data (Spamann 2015) in general, and to national courts judgments applying EU law provisions beyond the field of competition law in particular. Systematic content analysis of legal text is a unique legal empirical methodology, representing a transformation from an authority-based to a scientific-based methodology involving investigations into underreported legal, economic, and political effects of rules and the decision-making process (Brook 2022). Shifting the focus away from leading cases towards the day-to-day application of the law, systematic content analysis of legal text brings the rigor of social science to the study of law (Hall and Wright 2008).

While systematic content analysis emerged from the study of the American legal system, this article’s second contribution aims to share our experiences of applying this methodology in the context of the EU judicial federalist system. Our empirical-systematic exploration of the review role of the national courts of the EU Member States is unique in that it represents one of the first large-scale efforts to study multi-court, multi-country decision-making in a comparative manner (cf. Bricker 2021). Producing comparative data requires a particularly careful coding protocol, analysis, and interpretatiaon to ensure the validity of comparative treatment of that data (Spamann 2009, 2015). In addition to introducing the Database, this article, therefore, seeks to identify the methodological challenges of constructing these types of cross-country comparative databases and offer some potential ways to overcome the difficulties and limitations encountered.

The following Section 2 first (briefly) outlines the decentralized system for judicial review of EU competition law enforcement. Section 0 introduces the database, and Section 4 discusses the challenges involved in undertaking this project and suggests best practices for creating cross-country comparative databases in general, and EU-wide databases of national courts’ operations in particular, aiming to provide useful insights for future empirical studies in other fields of law.

2 Overview of Judicial Review of EU Competition Law Enforcement

Competition law (antitrust) enforcement in the EU is heavily reliant on public enforcement by the European Commission and the national competition authorities (NCAs) of the Member States. Unlike the US, where private antitrust enforcement is considerably more prominent than federal public enforcement by the Federal Trade Commission (FTC) and the Department of Justice (DoJ) (United States Courts 2024), private enforcement is still underdeveloped in Europe (Rodger et al 2023).¹⁰

Until May 2004, the public enforcement of the EU prohibitions against anti-competitive agreements (governed by Article 101 TFEU) and abuse of dominance (governed by Article 102 TFEU) was a centralized task undertaken by the European Commission (through its Directorate-General for Competition (DG Comp, earlier known as DGIV)). This old enforcement framework obliged firms to notify the European Commission about all agreements that had the potential to be considered as anti-competitive prior to their implementation, and the European Commission alone held the power to issue exemption decisions under Article 101(3) TFEU.¹¹ Consequently, the enforcement of the competition rules was virtually the sole responsibility of the European Commission and its decision-making was (and remains) subject to review by European judicial institutions: the GC at first instance, and the ECJ at second instance.

At the national level, the NCAs and the national courts of the EU Member States played only a limited role in relation to the enforcement of EU competition law. Although the ECJ had held that Articles 101(1) and 102 TFEU were directly applicable and could be enforced by the EU Member States’ NCAs and national courts, and despite the fact that around half of the NCAs were formally competent to apply Article 101(1) TFEU under their respective national laws, the European Commission’s monopoly in the application of Article 101(3) TFEU had in practice discouraged national enforcement. Alongside the EU competition law prohibitions, some of the Member States had also adopted national competition rules and created NCAs, especially from the 1990s onwards.¹² Nevertheless, with the exception of Germany, the enforcement of national competition laws was considerably limited.¹³

NCAs and national courts were awarded a more significant role at the start of the millennium, as EU Regulation 1/2003 (the ‘Regulation’) was introduced to facilitate the decentralization of the enforcement (and consequently the judicial review) of EU competition law, sharing the enforcement tasks with national actors.¹⁴ Since its entry into force in May 2004, around 90% of Articles 101 and 102 TFEU enforcement decisions have been rendered by NCAs,¹⁵ in addition to active enforcement of purely national cases in many Member States.

The Regulation has significantly transformed the role of national courts, which became “an important arm of application of the EU competition rules”.¹⁶ As illustrated by Figure 1, under the new enforcement regime, any decision by an NCA applying EU and/or national competition rules can only be reviewed by national courts. The EU Courts can only indirectly influence the decision-making of national courts by means of the preliminary reference procedure.¹⁷

The institutional structure of this decentralized judicial review system was highly contentious. During the long negotiations between the EU Member States and EU institutions leading to the adoption of the Regulation, not all parties agreed that a decentralized judicial review system — in addition to the decentralized enforcement system — was desirable. A number of stakeholders called for a centralized judicial review system, whereby the EU Courts would review appeals launched against both the decisions of the European Commission and also the NCAs applying EU competition law, warning that “[d]ivergent application of Community competition law by national courts would pose a threat to the proper functioning of the single market and the coherence of the system”.¹⁸ For example, the Economic and Social Committee, an important consultative body within the EU, suggested on various occasions that the judgments of national courts should be open to appeal to supranational courts (i.e., the EU Courts) empowered to assess both the law and the facts.¹⁹ According to the Committee, “in a decentralised system without a single appeal authority it is difficult to guarantee not only the right of defence, but also the coherent and consistent application of Community competition rules across the EU”.²⁰ Scholars, moreover, voiced concerns over the ability of the national courts to perform effective judicial review, noting that most national courts had only limited experience with the technical and highly specialized provisions of competition law (Ehlermann and Atanasiu 2002).

Nevertheless, the European Commission, which was the main driver of the reform process, disagreed. It maintained that as EU competition law and policy had been developed and clarified over the years leading to the reform, it was appropriate for the burden of enforcement to be shared “more equitably” with NCAs and national courts, “which have the advantage of proximity to citizens and the problems they face”.²¹ As a result, not only did the Regulation vest national courts with the onerous task of reviewing the vast majority of EU competition law enforcement, but it also contains very few provisions aiming to ensure the harmonized and consistent application of the EU competition rules across the internal market.

The EU’s decentralized competition law enforcement system is based on the broader EU law principles of procedural and institutional autonomy.²² Developed by the case law of the EU Courts, this notion entails that in the absence of specific EU rules, the Member States are free to determine their own procedural rules for implementing and enforcing the EU rules and to establish their own legal and administrative institutions for undertaking this task.²³ In particular, the Member States are allowed to adopt their own rules on the administration of justice based on their legal traditions (Halberstam 2021). This principle reflects the EU’s character as a ‘vertical’ federalist system, in which the administration of EU law mostly relies on the resources and institutions of its Member States. As part of this vertical system, the EU relies on national judiciaries to apply EU law to disputes in courts.²⁴

The principle of procedural and institutional autonomy is deeply embedded in Regulation 1/2003, setting out the decentralized competition law enforcement system. Accordingly, the institutional makeup, functions, and powers of the NCAs and national courts are primarily determined by the procedural enforcement powers and institutional settings established and prescribed by their respective domestic laws. EU law “affects neither national rules on the standard of proof nor obligations of competition authorities and courts of the Member States to ascertain the relevant facts of a case, provided that such rules and obligations are compatible with general principles of Community law”.²⁵

Regulation 1/2003 does not demand the involvement of any specific type of institution or procedure. In particular, the Regulation defines the notions of “national competition authority” and “national courts” in functional terms in recognition of the considerable institutional divergence across Member States. NCAs are “the authorities designated by the Member States including courts that exercise functions regarding the preparation and the adoption of the types of decisions” foreseen in the Regulation (emphasis added, Article 35(3). National courts, similarly, are the institutions “acting (…) as review courts” (emphasis added, Preamble 21). The Regulation does not indicate, for example, if the national courts should be constituted by a specialised tribunal, a generalist (civil, administrative, or criminal) court, or a specialised chamber of a generalist court. Similarly, it does not prescribe the number of instances of appeal, grounds of appeal, limitation periods, fees for launching appeals, which attributes the judges should have (e.g., specialised, ordinary judges, or layperson), nor the number of judges sitting on any panel. In the absence of any harmonising provisions, it was also our hypothesis that the standard, scope, and intensity of judicial review would also vary considerably across the EU and lead to different review processes and outcomes.

In 2004, the Commission introduced some soft measures to facilitate a degree of convergence within the European Competition Network (ECN),²⁶ which is an informal forum for discussion and cooperation between the European Commission and NCAs. The new enforcement regime resulted in voluntary harmonization between the Member States in some areas of competition law enforcement.²⁷ Later, the ECN+ Directive of 2019 marked a step towards greater harmonization, yet most of the Directive’s provisions are directed towards the functioning of the NCAs, while the autonomy of the national courts was left largely untouched. Article 3(2) of the ECN+ Directive merely provides a general obligation, according to which: “Member States shall ensure that the exercise of the NCA’s enforcement powers is subject to appropriate safeguards in respect of the undertakings’ rights of defence, including the right to be heard and the right to an effective remedy before a tribunal” (emphasis added).

Consequently, neither Regulation 1/2003 nor the ECN+ Directive sought to formally harmonize the domestic rules on enforcement procedures and the powers of the various NCAs and national courts. In particular, as a consequence of the principle of national procedural and institutional autonomy, there is a landscape of different court review systems in terms of the nature and degree of specialization of national courts, their respective roles and the scope and intensity of review available.²⁸

Figure 1. Institutional structure of EU competition law enforcement and judicial review

The same is true when it comes to the objectives of judicial review of EU competition law enforcement. Identifying the role for and scope of judicial oversight over the decisions of administrative law enforcement by independent and highly specialized administrative authorities has been subject to debate in scholarship for many years including but also beyond the field of competition law (Mantzari 2022). The Treaties, EU secondary legislation and soft laws, and the EU Courts’ jurisprudence provide limited guidance for the role and functioning of those national courts beyond general EU law principles.

To guide the discussion and assessment of the operation of the national courts, in this study, we differentiated between four possible tasks of national courts. The first three tasks are common to the review of the application of both the EU and national competition rules, while the fourth is limited to cases where EU competition law applies (that is, infringements having an effect on trade between Member States). These court tasks are as follows: (i) ensuring the appropriate interpretation of the substantive norms of the competition law provisions and their application; (ii) protecting the procedural due process and the fundamental rights of the parties involved or affected by the enforcement; (iii) verifying that the penalties imposed by the NCAs are procedurally and substantively effective, proportionate, and dissuasive; and (iv) ensuring that the NCAs apply the EU competition law provisions in a consistent and uniform manner across the EU internal market, and thereby preserve the effectiveness of EU law.

3 The EU Competition Law Judicial Review Database

Our main methodology for exploring the operation and effectiveness of the national judicial review of EU and national competition law enforcement was the construction of an original and open access database, covering all publicly available judicial review judgments of final public enforcement actions in relation to Articles 101 and 102 TFEU and the national equivalent provisions by NCAs. The database consists of panel data of 28 sub-datasets, for each of the EU’s current 27 Member States as well as the UK which was an EU Member State until the 31 of January 2020.²⁹

The construction of the database was led by the authors, with the assistance of a European-wide Advisory Board and twenty-eight national teams of competition law experts (the ‘coders’ or ‘national experts’). The experts are academics and practitioners, each with extensive knowledge of their respective competition law system.³⁰ This composition of our team was carefully selected to address the challenges posed by the diverging procedural and institutional frameworks of each national legal system. As detailed below, we utilized the Functional Comparative Approach to structure the coding and the analysis of the data, which is used to compare the structure, functions, and processes of different socio-political and legal systems.³¹ As elaborated in Section 3.6 below, this database is intended to facilitate academic and policy research into the operation and effectiveness of judicial review of competition law enforcement, from a comparative, EU-wide, and national perspectives. Given that the database is not designed to answer a specific research question, the variables were created to capture broad measures and indicators that could enrich multiple users (Weinshall and Epstein 2020).

3.1 Team and Coding Protocol

All variables were hand-coded. Coders followed a dedicated coding book to apply systematic content analysis of legal text (‘coding’) to the judgments.³² The full coding book is available on the project’s website.³³ It offers detailed coding instructions, aiming to ensure uniformity between the analysis of the jurisdictions.

The coding book was drafted by the project leaders, together with the Advisory Board.³⁴ Before launching the study, each member of the Advisory Board coded twenty judgments in an initial ‘pilot’ study, aiming to test the fit and accuracy of the coding book. This pilot study was performed across six jurisdictions (Hungary, Poland, Portugal, Spain, the Netherlands, and the UK), representing the EU Member States’ diverse legal traditions, history, and institutional settings. The coding book was subsequently revised, tweaked, and shared with the national experts. Next, to ensure inter-coder reliability, all national teams participated in online training, followed by a second ‘pilot’ exercise in which twenty judgments were coded by each team and reviewed by the Advisory Board. Small but necessary clarifications were added to the coding book and shared with all national teams. Questions arising during the coding process were answered by the editors based on the coding book and, when necessary, were consulted upon among the Advisory Board. When doubts arose, we tried to follow the functional comparative approach described below to guide the answer, interpreting the coding in light of Regulation 1/2003 rather than national law.

Based on the database, each national team of experts drafted a national report, providing an introduction to the national competition law enforcement framework, an outline of the national appeals system set in place to review the NCA’s enforcement decisions, and discussion of any prior studies which have explored the operation of judicial review in that jurisdiction, focusing in particular on any existing empirical studies. The national reports also provide a quantitative analysis presenting the empirical findings, and a more qualitative discussion, putting the empirical findings in context, followed by some concluding remarks.³⁵ Aiming to facilitate cross-jurisdictional comparison, the national reports follow an identical structure and make use of the same types of graphs and figures that were programmed by the project’s leaders. The reports are published in an edited collection (Rodger et al 2024).

3.2 Time Frame

The database includes all of the relevant national courts’ judgments that were issued and made public between 1 May 2004 and 30 April 2021. The starting date represents the entry into force of Regulation 1/2003, which decentralized the enforcement of competition law in the EU. The database does not include appeals on older NCAs’ decisions that were rendered prior to the decentralization of EU competition law. While this date was adopted to ensure uniformity, it should be noted that in some Member States, the Regulation was implemented at a later date. This timeframe was also used for those Member States which joined the EU after 2004 — namely Romania and Bulgaria (2007) and Croatia (2013) — and for those exiting the EU (the UK, 2020).

3.3 Case-level Variables

The database includes 32 case-level variables. Those variables provide quantitative and qualitative information about all judgments issued by the national courts in the EU Member States concerning the judicial review of public enforcement of Articles 101 and 102 TFEU and the national equivalent provisions by the NCAs. The units of analysis are the judgments of the courts.

As illustrated by Table 1, the database consists of all 5,707 judgments. The number of judgments is neither distributed equally across the Member States, nor according to their population or size of economy. For example, the figure reveals that there have been no appeals at all in Estonia and only two in Ireland, and that despite being the three largest economies in Europe, the courts in Germany, France, and the UK have issued only a relatively low number of competition law appeal judgments.

By comparison, the courts in Poland and Spain have issued the highest number of judgments during the relevant period. Our project suggests that the large number of appeals in Spain and Poland is not explained by a comparatively high ratio of appeals on such NCA’s decisions, but rather on national procedural and institutional particularities. Hence, in Poland, the high number of appeals is explained by the combination of the relatively high number of NCA decisions rendered during the relevant period that were potentially subject to an appeal and the frequent use of second- and third-instance appeals. In Spain, the large number of judgments is explained by the right of each infringing party to challenge any sanctioning decision with respect to its specific participation in the infringement in a separate proceeding. In one case,³⁶ a single NCA decision was subject to over 80 separate first- and second-instance judgments (Brook and Rodger 2024).

The database only includes judicial review of final NCAs’ decisions, including (i) infringement decisions, including in cases where the decision was adopted following a settlement with the NCA; (ii) decisions imposing fines for an infringement, either on a company or an administrative fine on individuals; (iii) no ground for action findings, concluding that the competition law prohibitions were not infringed; (iv) decisions to close or not to pursue a case (rejection of complaints, decisions not to investigate or stop investigation, including due to lack of evidence or prioritization); (v) NCAs’ decisions on recalculation of fines, adopted following an appeal; and (vi) decisions accepting commitments. As elaborated in Section 0 below, not all of those decisions are subject to appeal in all of the jurisdictions examined by this study.

The database does not include (i) judicial review/appeals of criminal or private enforcement cases; (ii) appeals on decisions of independent regional competition authorities; (iii) internal review proceedings of the NCA; and (iv) appeals on injunctions or monitoring and compliance orders adopted by NCAs. It also does not include judicial review of NCAs’ decisions and proceedings regarding interim measures; procedural matters (e.g., decisions on dawn raids, decisions regarding the right to be part of proceedings, confidentiality matters, fines for lack of cooperation or non-compliance); and decisions related to monitoring previously adopted NCA decisions. Likewise, the database does not include court judgments concerning an accepted application for leave for appeal³⁷ and cases that were paused and referred for a preliminary ruling on the interpretation of EU law under Article 267 TFEU. The latter would only be included in the database after the ECJ had provided its ruling and the national court had consequently delivered its judgment.

As summarized by Table 2, the case-level variables are organised in six sets: case identification, parties to the appeal, NCA’s procedure subject to an appeal, the anti-competitive behaviour, grounds of appeal, and outcome. In addition, the dataset includes an open-text comments field, allowing the coders to record any further information about the case.

The case identification variables, as the name suggests, provide information about the relevant jurisdiction, judgment date, relevant court, judge(s) presiding, judge rapporteur, docket number, and case name. As elaborated in Section 0 below, some of the Member States place legal restrictions on identifying the name of the judges presiding with the specific purpose of inhibiting forms of systematic research. In those jurisdictions, this information was not included in the database.³⁸ Similarly, some of the Member States (mostly those based on civil law traditions) make use of a judge rapporteur, a judge who furnishes a written report on the case at hand, in which he or she sets forth the arguments of the parties, specifies the questions of fact and of law raised in the dispute, as well as the available evidence.³⁹ Finally, some national judicial systems do not assign identifiable docket (case) numbers to judgments. In those instances, the coders have assigned their own case identifiers, as to allow for the ‘trail’ of appeals across instances to be followed.

The variables relating to the parties to the appeal involve information about the type of applicants in each judgment (e.g., the firm or individuals subject to the NCA’s decision, the NCA, or third parties), whether any other third parties have joined the appeal, and if so — their characteristics (e.g., the European Commission, a national government or agency, consumer organization, the complainant in front of the NCA, any firms or individuals that were part of the NCA’s decision but are not applicants/respondents in the appeal, or any other type of third parties).

The variables related to the NCA’s decision subject to an appeal encode information about the procedure subject to the appeal. They provide information about the case number of the original NCA’s decision subject to an appeal (or the previous instance’s judgment), the type of the NCA’s procedure (e.g., infringement decision with or without imposition of a fine, no grounds for action finding, decision accepting commitments or closing an investigation), the remedies imposed by the NCA (structural or behavioural), the level of fines imposed, and whether one or more of the parties to the original procedure have successfully applied for leniency⁴⁰ or a settlement.⁴¹

The anti-competitive behaviour set of variables is intended to control for the different types of anti-competitive conduct subject to an appeal. Those variables indicate the competition law prohibition(s) to which the original case related (e.g., anti-competitive agreements and/or abuse of dominance), whether the EU and/or national law provisions have been applied, the type of restriction of competition (e.g., horizontal or vertical anti-competitive agreements, exploitative or exclusionary abuses, and for the anti-competitive agreement cases, whether the NCA found the agreement to be a by-object or by-effect restriction of competition). Figure 2, provides an example of an anti-competitive behaviour variable, pointing to the percentage of first-instance judgments in each jurisdiction concerning the prohibition on anti-competitive agreements alone (Article 101 TFEU and/or the national equivalent). The Figure illustrates that in 16 of the 28 Member States, national courts have predominantly focused on the prohibition against anti-competitive agreements (more than 60% of the first-instance judgments, and in 6 Member States more than 80% of the judgments), and thus scarcely exercised judicial review on the application of the prohibition against abuse of dominance. By comparison, in a small minority of mostly Central and Eastern European Member States, the majority of appeals over the project period concerned the abuse of dominance prohibition. We argue that potential explanations for this focus on abuse of dominance cases in those jurisdictions include the economic transformation which has occurred in those Member States, the heavy concentration of markets at a local level due to the relatively small size of the various economies, and certain procedural rules that require NCAs to investigate all complaints (Rodger et al 2024).

The grounds of appeal variables indicate whether the appeal was deemed admissible, whether procedural or substantive grounds were raised and accepted, whether the fine was contested and reduced, whether the appeal involved questions about the tension between EU and national competition laws, and whether a preliminary reference to the ECJ had been submitted. The classification of grounds into those categories is particularly challenging in an EU-wide comparative setting, because some of the national legal systems classify similar grounds of appeal differently. To allow for a meaningful comparison across jurisdictions, the coding book included a detailed list of claims that should fall under each ground for the purpose of this study.

Finally, the outcome variables indicate the success of the appeal (e.g., rejected or fully/partially accepted), the outcome (e.g., full/partial annulment of the NCA’s decision or previous instance judgment, replacement of the NCA’s decision or previous instance judgment by that of the reviewing court, or an order returning the case to the NCA or previous instance court, or a combination of outcomes), and any change to the original fine imposed by the NCA.

Figure 2. Percentage of judgments concerning the prohibition on anti-competitive agreements alone.

3.4 Jurisdiction-level Variables

The database also includes 50 jurisdiction-level variables, providing background information about the relevant national competition law prohibitions equivalent to Articles 101 and 102 TFEU,⁴² and the currency used to impose fines. They also contain information on each of the relevant national courts (including, name, code, the level of the instance of appeal, whether the court is specialized in competition law matters, whether it can examine matters of facts and/or law, the standard for judicial review, whether leave for appeal is necessary, the limitation periods, the fee for submitting appeals, and whether there was a reform in the national judicial system during the relevant period). Finally, the jurisdiction-level variables detail the number of enforcement actions taken by the NCA during each of the relevant years covered by the project (both with including and exclusion of decisions rejecting complaints).

The jurisdiction-level variables clearly demonstrate the effects of the EU principle of procedural and institutional autonomy. For example, Figure 3 depicts the different number of possible instances of appeal that can be launched against an NCA decision applying the EU competition rules, ranging from a single instance of appeal in Austria, to four instances of appeal in Denmark. In other words, although all NCAs apply the same substantive EU competition law prohibition, the scope and nature of the judicial review of such decisions, and the protection afforded to the infringing parties’ rights, varies considerably across the EU.

Figure 3. Number of instances of appeal⁴³

3.5 Source of Information

For the case-level variables, the database gathered all published and publicly available judgments. This was a challenging endeavour, as there is no EU-wide public or private depository of the decisions of NCAs or of the national courts’ judgments. In many of the Member States, moreover, not all judgments are regularly or fully published (see discussion in Section 4.1 below). Each national sub-set, therefore, indicates the source of information about the national judgments, and whether its respective national database is expected to be comprehensive.

The coding of the case-level variables relied on information extracted from the relevant judgments. Some information related to unpublished or partially published judgments was retrieved from the NCA’s decision that was subject to the appeal or from subsequent appeals on the unpublished judgment. However, the coding did not record exogenous information from external sources, such as background information, news items, or scholarship.

The jurisdiction-level variables were coded based on the provisions of the national laws, and the considerable practical expertise of the national teams. They also reflect any procedural or institutional reforms taking place in the jurisdiction during the period covered by the database.

3.6 Facilitating Further Research Questions

This database is designed to support academic and policy research into the operation and effectiveness of national courts reviewing EU and national competition law enforcement. We believe it offers a valuable resource for exploring a wide range of research questions, including those focused on (i) (EU) competition law enforcement; (ii) the operation of civil procedure and courts; and (iii) empirical comparative analysis. The database can be utilized at both national and cross-country levels, and provide a foundation for descriptive, predictive, or prescriptive research purposes (McGill and Salyzyn 2021).

First, the database facilitates empirical investigation into long-standing competition law questions that have been subject to (doctrinal) legal research for many years. It can be used to analyse the success rates of different grounds of appeal, and the impact of the types of anti-competitive conduct and the EU and national provisions applied to those rates. Additionally, it may allow for analysis of the experience of various national courts in handling complex economic matters, the effects of EU and national competition law reforms (e.g., changes in the legal provisions, institutional structures, and procedural rules) and of significant legal precedents on judicial review. Moreover, the database provides an evidence-based basis for evaluating the impact and success of Regulation 1/2003 and the decentralized enforcement of EU competition law, particularly regarding judicial protection, the uniformity of application across the EU internal market, and the consistency and legal certainty for the business community through legal controls, including the role of preliminary references. Such evidence can be used to confirm or refute the assumptions made by the proponents of 1/2003 and whether the procedural and institutional autonomy principle meets the objectives of such regulation.

Second, the database also provides a basis for the exploration of issues related to civil procedure and the role of courts, contributing to scholarship on the courts’ central role in legal and political processes (Menkel-Meadow and Garth 2010). It can be used to study the effects of court expertise, interactions between highly specialized independent administrative regulators and the reviewing courts, and aspects of civil procedure, such as the length of the judicial review process, interactions between different tiers of appeal, and the impact of different types of appellants and third parties (e.g., complainants, government bodies, NGOs and CSOs). The database can be used to explore the function of judicial review in certain special procedures such as negotiated penalty settlements (leniency and settlements) and soft enforcement (commitments), and provide data for judicial analytical studies, investigating the behaviour and impact of specific courts, judges, or judge rapporteurs on appeal success rates and specific grounds of appeal.

Finally, the database supports potential comparative law projects that would examine if and how differences in processes and institutions (e.g., fora, decision-makers, procedures) can influence judicial review outcomes. Further research might explore the structure of national enforcement systems and their effects on the nature of judicial review and its outcomes (e.g., administrative v. judicial models, number of appeal instances, role of constitutional and internal reviews), standards and intensity of review, types of administrative decisions subject to review, and the influence of different legal traditions and cultures.

While many of those topics have been partially explored in our project’s national reports and comparative analysis, we anticipate that the database will enable further research by legal scholars and social scientists, thereby enriching the field with new insights and analyses.

4 Challenges for Creating a Multi-court Multi-country Database of National Judicial Review of EU Law

This section discusses the challenges we faced when creating the multi-court multi-country database of national judicial review of EU (competition) law, and shares our experience and good practices that might inform similar projects in the future. Such challenges include technical barriers common to many European jurisdictions, which limit the gathering and analysing of data necessary for conducting empirical legal research (e.g., limited publications of NCAs’ decisions and national courts’ judgments, weak open judicial data practices; legal restrictions imposed on the analysis of judicial data; and the various European languages), as well as the substantive hurdles affecting the design of a comparative-empirical study (considerable national divergence emerging from the different legal traditions of the Member States and the EU principle of procedural and institutional autonomy). The following sub-sections aim to contribute to the field of empirical comparative legal studies by discussing each of those challenges in turn and suggesting routes to overcome them.

4.1 Limited Publication of NCA’s Decisions and National Court Judgments

Full publication of judgments is essential for both the public scrutiny of judges and courts and for public knowledge and awareness of the development of law (Van Opijnen et al 2017; Huq and Clopton 2024). Naturally, it is also a basic requirement to facilitate an empirical-systematic examination of case law.

In the field of competition law enforcement, in particular, a complete database of decisions and judgments is essential for the promotion of both a better understanding and critical analysis of competition law enforcement (Posner 1970; Kovacic 2005). It is necessary to facilitate compliance by businesses with the open-ended provisions of EU competition law, and for scholarly and policy-driven research on the uniform application of the rules across the EU internal market. Full publication supports the interests of anyone harmed by an infringement so that they can assert their rights, as well as protecting the rights of any infringing parties seeking judicial review.⁴⁴

The importance of full publication of decisions and judgments is recognized at the EU level, namely with respect to the enforcement decisions of the European Commission and the EU Courts. Accordingly, the judgments of the EU Courts are generally published in accordance with the Rules of Procedure of the Court of Justice and the General Court,⁴⁵ and the European Commission’s competition law decisions are published as a rule, while non-disclosure is the exception (Iannuccelli and Nuijten 2023).

As a result, the EU Court’s judgments and the European Commission’s competition law decisions are relatively easy to gather systematically,⁴⁶ subject to noticeable reservations about the lack of automated tools to gather and analyze them (Ovádek 2021), lack of full-text search of the Commission’s decisions (Stylianou and Iacovides 2024), omissions of key documents that are rectified only following request for information notices (Van Rompuy 2022), and some long delays in publication of non-confidential versions of the decisions and judgments (Malnar 2015).

Access to national decisions and judgments applying EU law, nevertheless, is considerably more limited in many of the Member States. This is not only a matter of technical publication difficulties or a lack of attention, but also in some cases evidences a principled opposition to full publication of decisions and judgments.

At the EU level, EU law does not impose a publication requirement on all decisions of national administrative authorities or judgments of national courts applying EU law. At the national level, the laws and practices of the various Member States considerably diverge in terms of the scope and nature of publication. This was demonstrated by the comparative study of Van Opijnen et al (2017), showing that publication requirements in some Member States apply only to certain courts (often those of higher instances), and that even those Member States that have adopted legal or policy frameworks that require full or partial publication have not always complied with such rules.

The historical European resistance against a fully-fledged publication rule for judgments is reflected in the policy of the Council of Europe, which involves 46 countries across Europe with the goal of upholding human rights standards. Its 1995 recommendations on selecting, processing, presenting, and archiving of judgments in legal information retrieval systems does not provide for a high standard of full publication of judgments, but rather seeks to balance between on the ‘one hand, broad and comprehensive access to information on court decisions and, on the other hand, that the accumulation of useless information is avoided’.⁴⁷

To strike this balance, the decision whether or not to publish a judgment in Germany, for example, lies in the hands of the presiding judges, and German national law does not require administrative authorities or courts to publish their decisions and judgments, and there is no public register indicating which cases have been decided (Podszun and Overhoff 2024).⁴⁸

Our project highlights the significant challenges involved in creating systematic EU-wide databases on national enforcement and judicial review of EU law. These challenges arise from the absence of an EU-wide obligation to publish all NCAs’ decisions and national courts’ judgments implementing EU law, coupled with the limited publication rules at the national level. The lack of comprehensive and uniform publication, and its impact on efforts to compile consistent and complete data across Member States, have not been previously examined by scholarship. Our national teams of experts faced different degrees of difficulty in locating all of the relevant cases. In particular, during all or part of the relevant period, there was no obligation to publish all judgments concerning the public enforcement of competition law in the Czech Republic, Germany, Estonia, Greece, Hungary, Latvia, the Netherlands, Poland, Romania, and Slovenia, resulting in selective access to judgments (Rodger et al 2024).

Our study also shows that in some countries that follow a formal rule of full publication, moreover, these obligations were not always respected in practice. In Portugal, for example, although the NCA is obliged to publish all of the judicial decisions on appeals in relation to fines imposed by it, there are often significant delays in the publication of the judgments and not all judgments are being published in practice. There are indications that some of those delays and omissions are not unintentional. In 2023, for example, the Portuguese NCA failed to publish a judgment adopted almost a year earlier in the important Banking Cartel case. The NCA did not publish the judgment even after it was asked to, and the judgment was only published after a court proceeding on this matter was brought by a consumer association (Sousa Ferro 2024).

Notably, the limited publication of national judgments does not only hinder the public and academic assessment of the operation of the courts, but also the ability of the European Commission and other EU institutions to capture the functioning of the decentralized competition law regime. Article 15(2) of Regulation 1/2003 attempted to enact a process for transmitting national judgments to the European Commission, obliging the Member States to forward to the European Commission ‘without delay after the full written judgment is notified to the parties’ a copy of any written judgment by national courts deciding on the application of Articles 101 or Article 102 TFEU. This obligation, nevertheless, was not respected in practice. The European Commission reported that it ‘received very few national court judgments’,⁴⁹ and thus, it has only limited information on the application of EU competition law generally and on the scope and nature of the national judicial review of EU competition law enforcement in particular.

We believe that our database can offer a first step to remedy some of the challenges associated with the limited publication of NCAs’ decisions and national courts’ judgments in the field of competition law. The database is the first to offer an as-comprehensive-as-possible list of judgments reviewing the public enforcement of EU and national competition law, as well as valuable information on the details and outcomes of the case. The database links each judgment to the relevant NCA’s decision and to previous or subsequent appeals, and can facilitate future empirical research at national and EU-wide levels. At the same time, the database and national reports indicate whether each national sub-set is likely to be comprehensive, and how it may affect certain results (e.g., the ratio of appeals on the NCA’s decisions or the success rates).

4.2 Weak Open Judicial Data Practices

The limited publication of national courts’ judgments is only the first hurdle faced when conducting empirical legal research on the operation of European courts. The principle of open judicial data, also requires the following characteristics to be satisfied (Tauberer and Lessig 2007): that the data is complete (all public data is made available); primary (data is available as collected at the source, with the highest possible level of granularity and not in aggregate or modified forms); timely (data is made available as quickly as necessary to preserve the value of the data); accessible (data is available to the broadest range of users for the widest range of purposes); machine-processable (data is reasonably structured to allow automated processing); non-discriminatory (data is available to anyone, with no requirement of registration); non-proprietary (data is available in a format over which no entity has exclusive control); and license-free (data is not subject to any copyright, patent, trademark, or trade secret regulation although reasonable privacy, security, and privilege restrictions may be allowed).

In the past decade, Open Data initiatives have gained considerable traction, leading to publication of datasets from various legislative and executive branches of government world-wide. Nonetheless, while open judicial data is an integral part of open justice, progress in opening up judicial decision-making and creating datasets has been considerably slower in many jurisdictions (Marković and Gostojić 2020). In particular, many characteristics of the EU’s Member States’ judicial systems and processes stand in the way of compiling a complete and meaningful database of national judgments and inhibits the introduction of common open judicial data practices.

The first difficulty relates to the ease of access to judgments. Article 6(1) of the European Convention on Human Rights (ECHR), for example, simply states under the heading of the right to a fair trial that “[j]udgment shall be pronounced publicly”. This has been interpreted as allowing for a different range of publication standards (Ondřejová et al 2015). In France, until 2016, this obligation was understood as merely requiring that judgments will be accessible to anyone who has requested them, but not as imposing an obligation to publish them online or in another easily accessible form (Amaro 2024). Indeed, when creating our database, some of the national teams of experts only managed to collect the relevant judgments as a result of their own personal connections with the data providers (e.g., members of the NCAs or national courts). The existence of such obstructive and non-transparent practices does not only create additional costs but constitutes a clear restriction on open justice (McConville 2007), and is inimical to in-depth and critical assessment and appraisal of the functioning of courts which is necessary in any democratic society.

The second difficulty relates to the format of publication. The availability of judgments is not simply a matter of physical or online access. Empirical legal analysis requires information to be organized in a uniform and organised manner so that content is structured into meaningful elements that can be searchable and inform the choice of selection criteria. This functionality is missing in some of the Member States. In Greece, Latvia, and Lithuania, for example, the available databases containing competition law judgments are not reliably searchable given ineffective filtering criteria (Kalintiri and Nteka 2024; Jerņeva 2024; Malinauskaite 2024). Similarly, the publication of many judgments in pdf format alone and the lack of standardization of publication formats for national judgments, metadata, and case identifiers obstructs automated gathering and machine-assisted analysis of judicial data in Europe (Marković and Gostojić 2020).

The third difficulty relates to the type of information contained in the published judgments. In various Member States, judgments are published following considerable redactions or anonymization, hindering the possibility of following the ‘trail’ of appeals or to understand its outcome. In our project, for example, such redactions and anonymization made it difficult to link the NCAs’ decisions to the first-instance courts’ judgments and to subsequent instances of appeals in Hungary, Germany, Latvia, and Romania (Nagy 2024; Podszun and Overhoff 2024; Jerņeva 2024; Almășan and Bogrea 2024). Other states such as Poland, moreover, mostly published the operative part of the judgment, without the reasoning that can help to evaluate the court’s operation and analysis (Bernatt and Janik 2024).

Similarly, most European national courts (as well as the EU Courts) make decisions collectively and publish their judgments in consensus. They do not disclose the judges’ votes, and do not indicate whether there were any dissenting opinions of members of the courts and the content of those opinions (Bricker 2021). This reflects the long tradition of secrecy in judicial deliberation, characterizing civil law jurisdictions (Kelemen 2013). In most cases, there are no public records of the deliberative process, and neither the draft opinions nor private opinions of former judges are available. Commentators have already noted the hurdles that such practices and the civil law conceptualization of the role of judges have on the study of judge-level factors of judicial law-making in the EU and its Member States, in comparison with many similar studies conducted in the US (Bricker 2021).

The publication of judgments, and the availability of data within those judgments, is the cornerstone of empirical legal research into national courts’ application of EU law. Nevertheless, our project confirmed the existence of a gap between the conception of ‘open’ justice in legal and scientific meaning (Pah et al 2020). Whereas scientific practice is grounded on a commitment to sharing data and enabling others to replicate findings, many of the Member States adopt a legal conception of openness which is limited to a commitment to carrying out legal proceedings and public acts in a public space.

Creating EU-wide databases of national courts’ judgments applying EU law can help overcome some of the hurdles for empirical legal studies arising from the weak Open Judicial Practices in Europe. Our project aims to contribute to scholarship by observing open judicial data good principles, offering an as-comprehensive-as-possible database, collected on the basis of primary sources, and which is timely, accessible, machine-processable, non-discriminatory, open access, and free to use under the terms of CC BY-NC license.

4.3 Legal Restrictions on the Analysis of Judicial Data

The lack of full publication and truly open judicial data in many of the EU Member States can be attributed not only to legislative restrictions or technological challenges, but also to a more principled opposition to widespread use of judicial analytics. While judicial analytics are a common tool to monitor, explain, and predict judicial behaviour in the US (Stewart and Stuhmcke 2020), some of the EU Member States restrict the analysis of judicial decision-making even with respect to data contained in published judgments.

For example, in Romania, a broad interpretation of the EU’s General Data Protection Regulation (GDPR)⁵⁰ prohibits the identity of the judges deciding on specific matters from being revealed even when the judgment is published (Almășan and Bogrea 2024).⁵¹ In France, a 2019 law reform, aimed at increasing the transparency of judicial data, criminalised the use of personally identifiable data concerning judges or court clerks for the “purpose or result of evaluating, analyzing or predicting their actual or supposed professional practices”.⁵² While this prohibition was formally justified by the desire to avoid both undesirable pressure on the individuals undertaking judicial decision-making and strategic behaviour by litigants following the grant of free online access to higher courts’ judgments, it was also understood as being motivated by judges’ preferences to avoid scrutiny and accountability (McGill and Salyzyn 2021).

The above legal prohibitions on the analysis of judicial data had to be taken into account when constructing our database. Hence, the database does not include information on the identity of the judges in Romania and France, for example. Nonetheless, we believe that addressing the resistance to openness and transparency is essential for advancing empirical legal studies in Europe, and studies applying systematic content analysis of legal text can play a pivotal role in this regard.

4.4 Language Barriers and Legal Traditions

Linguistics clearly pose a hurdle to an EU-wide study. A significant strand of scholarship has been dedicated to the role of the multilingual environment of EU law and governance (Doczekalska 2018; Marino 2018). At the EU centralized level, all legal acts and measures must be drafted in the EU’s 24 official languages of the Member States,⁵³ yet some of the EU Courts’ judgments are excluded from these rules and are only published in the language of the case and the language of deliberation.⁵⁴ The ECJ stressed that EU provisions “must be interpreted and applied uniformly in the light of the versions existing in the other [EU] languages”. This principle implies that even provisions that appear clear and unambiguous in one language version may take on a different meaning when considered alongside other language versions.⁵⁵ At the national level, the implementation and application of EU law is mostly limited to each Member State’s official language(s). This entails that most of the application of EU law is not easily accessible to those who are not proficient in those national languages.

Effectively analysing the application of EU law by national courts, moreover, often presupposes not only knowledge of the relevant language, but also familiarity with the national legal terminology of the legal system (Mousourakis 2019). For example, difficulties in interpreting legal terms have arisen even in Member States having the same language (e.g., Switzerland, Austria and Germany, see Kauffmann 2013).

Considerably different legal traditions are a second barrier for a comparative-systematic study of national courts. The thirteen states originally establishing the United States of America, all had common-law traditions with a high degree of similarity. By comparison, the legal traditions of the EU Member States differ significantly. In addition to the English common-law approach, some Member States follow the French or German traditions, there are various mixed-systems, and some States are deeply influenced by the communist traditions of their past (Koopmans 1991; Mousourakis 2019). The diverging legal traditions are an expression of the “nature and behaviour of traditions in social life” (Krygier 1986). Such legal traditions affect the historical background and development of each system, its predominant and characteristic mode of thought in legal matters, its distinctive institutions, the kind of legal sources it acknowledges and the way it handles them, and its ideology (Zweigert and Kötz 1987).

The interplay between the rich multilingual character of EU law and the diverse legal traditions within its Member States fosters the concept of 'plurilingualism.' This notion underscores the critical importance of being aware of the dynamic use of multiple languages, as well as the cultural knowledge and experiences that shape social interactions (Husa 2022). We strived to mediate the difficulties associated with the various languages and legal traditions by relying on the EU-wide teams of experts, each having a deep knowledge of the law, history, and traditions of their relevant Member States (cf Spamann 2015). While such “level of knowledge that is very difficult, if not impossible, for a single scholar to attain”, using national experts to construct the database (Mousourakis 2019), as well as an EU-wide representative Advisory Board to structure the project, define the variables, and examine the pilot coding, can help to mediate some of those difficulties. In addition, the national reports drafted alongside the database, explores the law, history, and institutions of each legal system, and flag any national particularities that should be taken into account when using and analysing the database.

4.5 The EU Principle of Procedural and Institutional Autonomy and the Functional Comparative Approach

Another key challenge surrounding the systematic study of national applications of EU law stems from the EU principle of procedural and institutional autonomy, leaving the Member States the power to determine their own procedural rules for implementing and enforcing EU rules and also to establish their own legal and administrative institutions. As discussed in Section 2 above, EU law refrains from determining the attributes of national courts, and the term ‘national courts’ acquires meaning only by reference to the function of a court as the institution reviewing an NCA’s public enforcement actions. This choice is striking especially as several Member States requested the adoption of a clear definition of the term during the negotiation procedure of Regulation 1/2003⁵⁶ and noted that the distinction between NCAs and courts in their judicial system is “problematic”.⁵⁷

Our project demonstrates that the divergences resulting from the principle of procedural and institutional autonomy create many challenges for an empirical exploration of national practices, and require tailoring the research design and definition of the selection criteria for the database:

First, the divergences resulting from the principle of procedural and institutional autonomy is evident when defining the relevant judicial institutions to be included in the study. Most of the Member States have adopted an ‘administrative enforcement system’ for competition law, in which an administrative competition authority acts both as the investigator and the first-level decision-maker. For those types of jurisdictions, the database includes all judicial review proceedings launched against the NCAs’ decisions and examined by courts or tribunals. Yet, some of the Member States opted for a ‘judicial enforcement system’.⁵⁸ In this institutional setting, which also characterizes US public antitrust enforcement, decisions on the application of the competition rules are formally adopted via a court or tribunal judgment, following a procedure (which varies from one Member State to another) initiated by the administrative authority. As a reflection of the principle of procedural and institutional autonomy, Regulation 1/2003 states that in the countries that follow the judicial enforcement system, the NCA should be the court or tribunal adopting the final decision, rather than the administrative competition authority.

The functional definition of NCAs in the Regulation informed our research design. Accordingly, for the purpose of this project NCA is defined as the first-level decision-maker(s) on matters related to the application of Articles 101 and 102 TFEU and the national equivalent prohibitions in public enforcement settings. It may be an administrative or a judicial body or a combination of both.

To complicate matters further, in some of the jurisdictions following the judicial enforcement systems model, not all competition law decisions are taken by a court or a tribunal. Decisions relating to the rejection of complaints, no grounds for action findings, and acceptance of commitments and/or settlements are often dealt with by the administrative authority without any need for the court’s authorisation/approval or decision. That means that in certain contexts and for certain processes and decisions, the administrative authority would be considered as an NCA, while in others, it would be the court or tribunal.

Moreover, some Member States have enacted constitutional types of review, outside of the regular appeals system, focusing on the protection of fundamental rights and the rule of law. Some of those constitutional reviews assess the NCAs’ final decisions themselves (e.g., Portugal and Malta), while others only focus on claims concerning infringement of constitutional rights and liberties during the process of adopting the NCA’s decision or the ordinary judicial review judgment (Croatia, Slovenia, and Spain) (Brook and Rodger 2024).⁵⁹ Other Member States have introduced an internal appeal-like review process, in which the decision of the NCA is re-examined within the NCA hierarchy, before an appeal can be submitted to a court (e.g., Czech Republic, Slovakia, and the Netherlands, see Brook and Rodger 2024).

Second, the divergence resulting from the principle of procedural and institutional autonomy affected the types of procedures, scope, and the intensity of the judicial review process. In this project, we used the terms ‘judicial review’ and ‘appeals’ interchangeably, as referring to the evaluation of the NCA decision by review courts in the meaning of Regulation 1/2003. This terminology, however, is not fully in line with certain legal traditions (e.g., the UK), where a distinction is made between ‘judicial review,’ which is a particular form of ‘limited’ review of the lawfulness of the decisions of public bodies, and ‘appeal’ which signifies a full appeal ‘on the merits’ normally from an earlier decision by a court (Mantzari 2022).

We encountered a similar challenge in the empirical-comparative study of judicial review pertaining to the type of NCAs’ decisions that could be subject to an appeal. Our project pointed to the considerable divergence across the Member States as to the types of (i) decisions each NCA can adopt; and (ii) the NCAs’ decisions that can be subject to review by national courts. In particular, only some NCAs are entrusted to adopt decisions accepting commitments or settlements, terminating an investigation, or rejecting complaints, and such decisions are subject to an appeal only in some of the Member States. Considerable national divergence also arises in terms of the type of parties that have standing in proceedings before the NCA and the national courts.

Finally, beyond questions of definitions and selection criteria, the above differences impact the calculations of common performance measures for evaluating the operations of NCAs and courts, such as the rate of appeals and the success rates. In this project, we decided to calculate those measures by relying on the total number of NCA’s decisions that could be subject to an appeal according to the national rules of each jurisdiction. However, based on the research question investigated, one could also compare the rate of appeal or the success rates across each type of NCA decision (e.g., infringement decisions, decisions to accept commitments etc.).

To overcome the difficulties associated with the diverging procedural and institutional setting of each national legal system, we structured the coding and the analysis of the data based on the Functional Comparative Approach (Michaels 2019; Mousourakis 2019). This approach is used to compare the structure, functions, and processes of different socio-political and legal systems, and is premised on the notion that different legal systems may adopt different legal measures to solve similar legal issues. Accordingly, the research methodology and the definition of the variables identified the issues in purely functional terms, without giving deference to the design of certain legal systems and legal cultures. Those variables seek to encapsulate the ways the different legal systems deal with and resolve the legal issues, and compare and evaluate their approaches and solutions from the lens of their functionality in dealing with the issue.

We believe that the Functional Comparative Approach is essential for studying multi-level governance networks such as those characterizing EU competition law enforcement, in which the identical EU provisions (Articles 101 and 102 TFEU) are being applied by NCAs and national courts in the context of diverging national procedural rules and institutional settings (Cseres 2010). This is reflected by the approach of the EU competition rules themselves, which are functional to the extent that they do not only focus on the form of the written EU and national rules but on their effects; “not on doctrinal structures and arguments alone but on the consequences they bring about” (Michaels 2019). The operation of the different NCAs and national courts and their national procedural rules and institutional designs are deemed to be comparable even if they are doctrinally different because they are functionally equivalent, that is, because they fulfil similar roles across the EU Member States’ legal systems.

This functional comparative approach, naturally, comes with a cost. Drawing a comparison between a large sample of significantly diverging systems requires a degree of generalization and simplification. In particular, using the same metrics and figures to interpret the national data means that those metrics and figures do not always perfectly fit national procedures, institutions, or terminology (Hadfield 2009; Spamann 2009). These challenges should inform the analysis of the database and the interpretation of the results.

At the same time, we believe that the functional comparative research design for the study of the decentralized application of EU law and judicial federalism can be defended on normative grounds, not only practical ones. The functional comparative approach is not only necessary to allow for robust comparison but is also supported by the fact that EU law relies on those considerably divergent national systems to perform a shared task.

5 Concluding Remarks

To date, there has only been limited empirical study of the operation and impact of national courts applying EU law, especially from an EU-wide comparative perspective. Although national courts are the main judicial forum applying EU law, most of the scholarly and policy attention has been dedicated to the operation of the EU Courts or using traditional doctrinal legal methods to assess the national courts. The substantial growth in the use of systematic analysis of courts judgments that is evident in North America has not yet translated to academic practice in Europe.

We believe that a comparative-systematic analysis of court judgments is particularly appropriate and necessary in the EU’s vertical federal setting, in which national agents (authorities and courts) are responsible for the vast majority of the application of EU law and its judicial review. While EU law expresses a strong commitment to uniformity and consistency of substantive rules across the EU single market, much of the enforcement of those rules is decentralized, relying on (considerably divergent) national procedures and institutions. This type of empirical study can provide much needed insights in terms of the actual functioning of the different national practices, allowing researchers to observe trends and inform policy.

This is not an easy task. In this article, we discussed the various unique challenges that stand in the way of empirical comparative analysis in general, and for the creation of any multi-country multi-court database of the national judicial review of EU law enforcement in particular, using our construction of the EU Competition Law Judicial Review Database as a case study. We pointed to technical barriers common to many European jurisdictions, which limit the gathering and analysis of data necessary for conducting empirical legal research, as well as to the substantive hurdles affecting the design of a comparative-empirical study. Nonetheless, we believe that using national teams of experts, an EU-wide advisory board, and the Functional Comparative Approach to guide the design and construction of such projects may overcome many of those challenges.

Acknowledgments

This database, and the underlying empirical research project, would not have come to life without the incredible hard work of all national teams of experts in compiling case databases, coding all relevant cases in conformity with our lengthy coding book, and writing their national reports following a structured template. This has taken a considerable period to finalise, and was fully undertaken on a voluntary basis. We are extremely grateful to all of them for engaging with this research methodology, including Adriana Almasan, Adriani Kalintiri, Alexandr Svetlicinii, Ana Vlahek, Anton Dinev, Caroline Cauffman, Caroline Heide-Jørgensen, Csongor István Nagy, Evelin Pärn-Lee, Jasminka Pecotic Kaufman, Julija Jerneva, Jurgita Malinauskaite, Lars Henriksson, Lefkothea Nteka, Maciej Janik, Maria Vassiliou, Marios Iacovides, Mary-Catherine Lucey, Michal Petr, Michele Messina, Nils Overhoff, Ondrej Blažo, Petri Kuoppamäki, Rafael Amaro, Rupprecht Podszun, Ștefan Bogrea, Sylvann Aquilina Zahra, and Viktoria Robertson. Moreover, we extend our gratitude to our fellow co-editors and Advisory Board Members — Annalies Outhuijse, Francisco Marcos, Maciej Bernatt, and Miguel Sousa Ferro — that all made this momentous endeavour a reality.

We also would like to thank Andy Gavil, Eun Hye Kim, and Simon Halliday, for their valuable comments and suggestions on this article.

This study was conducted without external funding, and the authors declare no conflicts of interest. As the research did not involve human or animal subjects or sensitive data, no ethics approval was required. All relevant data supporting the findings of this study are available on our website: https://www.mappingcomplawreview.com. For any additional inquiries, please contact the corresponding author.

References

Alemanno, Alberto, and Oana Stefan (2014) "Openness at the Court of Justice of the European Union: Toppling a taboo" 51(1) Common Market Law Review. https://doi.org/10.54648/cola2014004.

Almăşan, Adriana, and Ştefan Bogrea (2024) “Romania Report,” in B. Rodger and O. Brook et al eds., Judicial Review of Competition Law Enforcement in the EU Member States and the UK. Kluwer.

Amaro, Rafel (2024) “France Report” in B. Rodger and O. Brook et al eds., Judicial Review of Competition Law Enforcement in the EU Member States and the UK. Kluwer.

Bernatt, Maciej (2012) “Convergence of Procedural Standards in the European Competition Proceedings” 8(3) Competition Law Review 255. https://doi.org/10.5235/20444153.8.3.25.

Bernatt, Maciej (2019) “Rule of Law Crisis, Judiciary and Competition Law” 46(4) Legal Issues of Economic Integration 345. https://doi.org/10.54648/leie2019022. 

Bernatt, Maciej and Maciej Janik (2024) “Poland Report” in B. Rodger and O. Brook et al eds., Judicial Review of Competition Law Enforcement in the EU Member States and the UK. Kluwer.

Bobek, Michal (2019), “Data protection, anonymity and courts” 26(2) Maastricht Journal of European and Comparative Law 183-189. https://doi.org/10.1177/1023263x19851628.

Bradford, Anu, Adam S. Chilton, Christopher Megaw, and Nathaniel Sokol (2019) "Competition law gone global: introducing the comparative competition law and enforcement datasets" 16(2) Journal of Empirical Legal Studies 411. https://doi.org/10.1111/jels.12215.

Bricker, Benjamin (2021) “Consensus decision making: A comparative analysis of judging and judicial deliberations” 42(1) Justice System Journal 25. https://doi.org/10.1080/0098261x.2020.1856738.

Brook, Or (2022). “Politics of coding: on systematic content analysis of legal text” in Bartl, Marija, and Jessica C. Lawrence, eds., The Politics of European Legal Research: Behind the Method. Edward Elgar Publishing. https://doi.org/10.4337/9781802201192.00015.

Brook, Or, and Katalin J. Cseres (2024) "Priority Setting as the Blind Spot of Administrative Law Enforcement: A Theoretical, Conceptual, and Empirical Study of Competition Authorities in Europe." 87(5) The Modern Law Review 1209. https://doi.org/10.1111/1468-2230.12881.

Brown, Rachel, Jade Ford, Sahrula Kubie, Katrin Marquez, Bennett Ostdiek, and Abbe R. Gluck. (2021) "Is unpublished unequal? An empirical examination of the 87% nonpublication rate in federal appeals." 107 Cornell Law Review 1.

Cseres, Katalin J (2010) ‘Comparing Laws in the Enforcement of EU and National Competition Laws’ 3 European. Journal of Legal Studies 7.

Doczekalska, Agnieszka (2018) "Legal languages in contact: EU legislative drafting and its consequences for judicial interpretation" in S. Marino, L. Biel, M. Bajčić, and V. Sosoni, eds Language and Law: The Role of Language and Translation in EU Competition Law. Springer. https://doi.org/10.1007/978-3-319-90905-9_10.

Dyevre, Arthur, Wessel Wijtvliet, and Nicolas Lampach (2019) "The future of European legal scholarship: empirical jurisprudence" 26(3) Maastricht Journal of European and Comparative Law 348. https://doi.org/10.1177/1023263x19840263.

Ehlermann, Claus-Dieter and Isabela Atanasiu (2002) “The Modernisation of EC Antitrust Law: Consequences for the Future Role and Function of the EC Courts” 23(2) European Competition Law Review 72. https://doi.org/10.5135/eusj1997.2002.332.

Emin, Charlotte (2023) “Luxembourg Confidential: Keeping (Business) Secrets Before the General Court of the European Union” 14(1) Journal of European Competition Law and Practice 4. https://doi.org/10.1093/jeclap/lpac048.

Geradin, Damien and Nicolas Petit (2012) “Judicial Review in European Union Competition Law: A Quantitative and Qualitative Assessment” in M. Merola and J. Derenne, eds., The Role of the Court of Justice of the European Union in Competition Law Cases. Bruylant.

Hadfield, Gillian (2009) "The Strategy of Methodology: The Virtues of Being Reductionist for Comparative Law" 59(2) University of Toronto Law Journal 223. https://doi.org/10.1353/tlj.0.0018.

Halberstam, Daniel (2021) “Understanding national remedies and the principle of national procedural autonomy: A constitutional approach” 23 Cambridge Yearbook of European Legal Studies 128. https://doi.org/10.1017/cel.2021.12.

Halberstam, Daniel (2008) “Federalism and the Role of the Judiciary’ in K. Whittington et al, eds, The Oxford Handbook of Law and Politics. Oxford University Press.

Hall, Mark A., and Ronald F. Wright (2008) “Systematic content analysis of judicial opinions” 96 California Law Review 63.

Hamann, Hanjo (2019) “The German federal courts dataset 1950–2019: From paper archives to linked open data” 16(3) Journal of Empirical Legal Studies 671. https://doi.org/10.1111/jels.12230.

Hesselink, Martijn W. (2002) The New European Legal Culture. Kluwer Law International.

Huq, Aziz Z. and Zachary D. Clopton (2024) “The Necessary and Proper Stewardship of Judicial Data” 76 Stanford Law Review 893.

Husa, Jaakko (2022) Interdisciplinary Comparative Law: Rubbing Shoulders with the Neighbours or Standing Alone in a Crowd. Edward Elgar Publishing.

Iannuccelli, P. and J. Nuijten (2023) “Article 30 Publication of Decisions: Commentary” in L. Prete, J. Laitenberger, C. Gauer, and K. Dekeyser, eds, Regulation 1/2003 and EU Antitrust Enforcement. Wolters Kluwer. https://doi.org/10.1007/s10991-023-09350-w.

Jerņeva, Jūlija (2024) “Latvia Report” in B. Rodger and O. Brook et al eds., Judicial Review of Competition Law Enforcement in the EU Member States and the UK. Kluwer.

Kalintiri, Andriani and Lefkothea Nteka (2024) “Greece Report” in B. Rodger and O. Brook et al eds., Judicial Review of Competition Law Enforcement in the EU Member States and the UK. Kluwer.

Kauffmann, Philipp (2013) “Spreading the Law – Comparative Legal Traditions” 2/2013 Journal on European History of Law 33.

Kelemen, Katalin (2013) “Dissenting opinions in constitutional courts” 14(8) German Law Journal 1345. https://doi.org/10.1017/s2071832200002297.

Kokkoris, Ioannis and Ioannis Lianos (2009) The Reform of EC Competition Law: New Challenges. Wolters Kluwer.

Koopmans, Thijmen (1991) “The birth of European law at the crossroads of legal traditions” 39(3) American Journal of Comparative Law 493. https://doi.org/10.2307/840768.

Kovacic, William E. (2005) “Using Ex Post Evaluations to Improve the Performance of Competition Policy Authorities” 31 Journal of Corporate Law 503.

Krygier, Martin (1986) “Law as Tradition” 5(2) Law and Philosophy 237. https://doi.org/10.1007/bf00190762.

Laborde, Jean François (2021) “Cartel Damages Claims in Europe: How Courts Have Assessed Overcharges” 3-2021 Concurrences 232.

Malinauskaite, Jurgita (2024) “Lithuania Report” in B. Rodger and O. Brook et al eds., Judicial Review of Competition Law Enforcement in the EU Member States and the UK. Kluwer.

Malnar, Vlatka Butorac (2015) “Access to documents in antitrust litigation–EU and Croatian perspective” 8(12) Yearbook of Antitrust and Regulatory Studies (YARS) 127. https://doi.org/10.7172/1689-9024.yars.2015.8.12.6.

Mantzari, Despoina (2022) Courts, Regulators, and the Scrutiny of Economic Evidence. Oxford Univ. Press. https://doi.org/10.1093/oso/9780198851608.001.0001.

Marcos, Francisco (2024) “Spain Report” in B. Rodger and O. Brook et al eds., Judicial Review of Competition Law Enforcement in the EU Member States and the UK. Kluwer.

Marino, Silvia (2018) “Introduction: The Role of the Language in EU Law” in S. Marino, L. Biel, M. Bajčić, and V. Sosoni, eds, Language and Law: The Role of Language and Translation in EU Competition Law. Springer. https://doi.org/10.1007/978-3-319-90905-9_1.

Marković, Marko and Stevan Gostojić (2020) “Open judicial data: a comparative analysis” 38(3) Social Science Computer Review 295. https://doi.org/10.1177/0894439318770744.

Martineau, Robert J (1994) “Restrictions on Publication and Citation of Judicial Opinions: A Reassessment” 28 University of Michigan Journal of Law Reform 119.

McConville, Michael (2007) “Development of empirical techniques and theory” In M. McConville, eds, Research methods for law. Edinburgh University Press. https://doi.org/10.1515/9781474404259-015.

McGill, Jena and Amy Salyzyn (2021) “Judging by Numbers- How Will Judicial Analytics Impact the Justice System and its Stakeholders” 44 Dalhousie Law Journal 249.

Menkel-Meadow, Carrie J., and Bryant G. Garth (2010) "Process, People, Power and Policy: Empirical Studies of Civil Procedure and Courts" in P. Cane and H. Kritzer, eds., Oxford Handbook of Empirical Legal Research. Oxford University Press.

Michaels, Ralf (2019) “The Functional Method of Comparative Law”, in M. Reimann and R. Zimmermann, eds., The Oxford Handbook of Comparative Law. Oxford University Press.

Mousourakis, George (2019) Comparative Law and Legal Traditions. Springer International Publishing. https://doi.org/10.1007/978-3-030-28281-3.

Nagy, Csongor (2024) “Hungary Report” in B. Rodger and O. Brook et al eds., Judicial Review of Competition Law Enforcement in the EU Member States and the UK. Kluwer Law International.

Ondřejová, Eva et al (2015) “Report on the Accessibility to Judicial Decisions through Publication Standards’ report prepared for the European Union and the Council of Europe”. Available at: https://rm.coe.int/1680700781.

Ovádek, Michal (2021) “Facilitating access to data on European Union laws” 3(1) Political Research Exchange. https://doi.org/10.1080/2474736x.2020.1870150.

Pah, A. R., Schwartz, D. L., Sanga, S., Clopton, Z. D., DiCola, P., Mersey, R. D. and Amaral, L. A. N. (2020) “How to build a more open justice system” Science, 369. https://doi.org/10.1126/science.aba6914.

Podszun, Rupprecht and Nils Overhoff (2024) “Germany Report” in B. Rodger and O. Brook et al eds., Judicial Review of Competition Law Enforcement in the EU Member States and the UK. Kluwer Law International.

Posner, Richard A. (1970) “A Statistical Study of Antitrust Enforcement” 13(2) The Journal of Law and Economics 365. https://doi.org/10.1086/466698.

Rodger, Barry J, Brook, Or, Bernatt, Maciej, Marcos, Francisco, and Outhuijse, Annalies (2024), Judicial Review of Competition Law Enforcement in the EU Member States and the UK. Kluwer Law International.

Rodger, Barry J, Miguel Sousa Ferro and Francisco Marcos (2023) Research Handbook on Private Enforcement of Competition Law in the EU. Edward Elgar.

Šadl, Urška, Lucía López Zurita, Irene Otero, Eun Hye Kim, och Stein Arne Brekke (2024) ”The IUROPA CJEU Database: Doctrine and Legal Outcomes” In The IUROPA Court of Justice of the European Union (CJEU) Database, (eds) Johan Lindholm, Daniel Naurin, Urška Šadl, Anna Wallerman Ghavanini, Stein Arne Brekke, Joshua Fjelstul, Silje Synnøve Lyder Hermansen, m.fl. https://www.iuropa.pol.gu.se.

Sauter, Wolf (2016) Coherence in EU Competition Law Oxford Univ. Press. https://doi.org/10.1093/acprof:oso/9780198749158.001.0001.

Sousa Ferro, Miguel (2024), “Portugal Report” in B. Rodger and O. Brook et al eds., Judicial Review of Competition Law Enforcement in the EU Member States and the UK. Kluwer.

Spamann, Holger (2009) “Large-sample, Quantitative Research Designs for Comparative Law?” 57(4) The American Journal of Comparative Law 797. https://doi.org/10.5131/ajcl.2008.0023.

Spamann, Holger (2015) “Empirical Comparative Law” 11(1) Annual Review of Law and Social Science 131. https://doi.org/10.1146/annurev-lawsocsci-110413-030807.

Stewart, Pamela and Anita Stuhmcke (2020) “Judicial analytics and Australian courts: A call for national ethical guidelines” 45(2) Alternative Law Journal 82. https://doi.org/10.1177/1037969x19899674.

Stone-Sweet, Alec and Thomas Brunell (1998) “The European Court and the National Courts: A Statistical Analysis of Preliminary References, 1961-95” 5 Journal of European Public Policy 66. https://doi.org/10.1080/13501768880000041.

Stone-Sweet, Alec and Thomas Brunell (2001) “The European Court, National Judges and Legal Integration: A Researcher’s Guide to the Data Base on Preliminary References in European Law, 1958-98” 6 European Law Journal 117. https://doi.org/10.1111/1468-0386.00100.

Stylianou, Konstantinos and Marios Iacovides (2024) “The DB-COMP database”, https://db-comp.eu/about-this-site.

Tauberer, Joshua and Larry Lessig (2007) “The 8 principles of open data government”, http://opengov data.org/.

Temple Lang, John (2014) “The Duty of Cooperation of National Courts in EU Competition Law” 17 Irish Journal European Law 27.

United States Courts (2024) “U.S. District Courts – Civil Cases Commenced, by Basis of Jurisdiction and Nature of Suit, During the 12-Month Periods Ending September 30, 2021 and 2022”. https://www.uscourts.gov/sites/default/files/data_tables/jb_c2_0930.2023.pdf.

Van Gestel, Rob, and Hans‐Wolfgang (2014). “Why Methods Matter in European Legal Scholarship” 20(3) European Law Journal 292. https://doi.org/10.1111/eulj.12049.

Van Opijnen, Marc (2016) "Court decisions on the internet: Development of a legal framework in Europe" 24(2) Journal of Law, Information and Science 26. https://doi.org/10.2139/ssrn.2794167.

Van Opijnen, Marc et al. (2017) “On-line publication of court decisions in the EU: report of the policy group of the project ‘Building on the European Case Law Identifier’”. https://doi.org/10.2139/ssrn.3088495.

Van Rompuy, Ben (2022) “The European Commission’s Handling of Non-priority Antitrust Complaints: An Empirical Assessment” 45(2) World Competition 265. https://doi.org/10.54648/woco2022010.

Vesterdorf, Bo (2005) “Judicial Review in EC Competition Law: Reflections on the Role of the Community Courts in the EC System of Competition Law Enforcement” 1 Global Competition Policy 1.

Weinshall, Karen and Lee Epstein (2020) “Developing High‐Quality Data Infrastructure for Legal Analytics: Introducing the Israeli Supreme Court Database” 17(2) Journal of Empirical Legal Studies 416. https://doi.org/10.1111/jels.12250.

Wells, Michael L. (2016) “Judicial federalism in the European Union” 54 Houston Law Review 697.

Zekoll, Joachim (1991) “European Community competition law and national competition laws: compatibility problems from a German perspective” 24 Vanderbilt Journal of Transnational Law 75.

Zglinski, Jan (2023) “The new judicial federalism: the evolving relationship between EU and Member State courts” 2(2) European Law Open 345. https://doi.org/10.1017/elo.2023.27.

Zweigert, Konrad and Hein Kötz (1987). An introduction to comparative law. Clarendon Press.