Who shapes the CJEU regulatory jurisprudence? On the epistemic power of economic actors and ways to counter it

Abstract

What role does the Court of Justice of the European Union (CJEU) play in holding European Union (EU) regulatory science, ie the science underpinning public regulation, to account for its epistemic quality? Since Pfizer, CJEU case law has strengthened the role of science in both EU risk regulation and litigation, whilst intensifying judicial scrutiny of scientific reasoning based on procedural standards. Scholars often welcome this approach as striking an adequate balance between effective judicial protection and institutional competence. Some praise the Court as an ‘information catalyst’ promoting procedures, in which epistemic quality is ensured through diligent consideration of and deliberation among all relevant voices. These debates, however, overlook the role of economic actors as information providers, in and challengers of EU risk regulation. This paper re-evaluates the modern post-Pfizer approach from a new, socio-legal perspective by studying, for the first time, the interactions between judicial review and the epistemic power of economic actors, ie their relative ability to influence what EU regulators know at the expense of other actors. We combine an epistemologically informed comparative institutional analysis with doctrinal critique of CJEU case law. Our findings show the need to rethink both legal standing and procedural review in EU risk regulation. Instead of catalysing inclusive procedures that open regulatory science to public scrutiny, the modern approach fosters an exclusive bilateral information exchange between the administration and the regulated industry. The Court reduces the function of process values, such as duty of care and reason-giving, to the protection of a small circle of actors, neglecting the public interest dimension of such values. Thus, the modern approach fails to address, and instead further entrenches the epistemic power imbalances inherent in EU risk regulation. We end by sketching out a normative and doctrinally sound vision of how CJEU review could contribute to EU expert accountability in a more publicly oriented way.

1. Introduction

Scientific advice plays a crucial role in risk regulation. As the current health and environmental crises demonstrate, no modern regulatory system can achieve its goals without it. At the same time, the problem-solving potential of public regulation critically hinges upon the epistemic quality of scientific advice. European Union (EU) regulatory science – ie the science used and produced for regulatory purposes¹ – is held to the high standards of independence, transparency, and excellence as formulated by the EU General Court in Pfizer.² Scholars and policy actors alike widely recognise that accountability mechanisms must reach beyond the final, political decision-makers, to the expert advisors who act backstage, yet often de-facto pre-determine final decisions on risk and technology.³ European courts play an important role in holding experts to account by probing the scientific foundations of regulatory decisions as part of their legality review. Over roughly the last two decades, the Court of Justice of the European Union (CJEU) has both acknowledged the importance of science in regulation and intensified judicial scrutiny of its quality. Scholars have praised this line of jurisprudence as overall a welcome response to the realities of a maturing EU risk regulatory state. Some have applauded the Court for upholding individual rights and ensuring judicial protection even against highly complex regulations. Others have celebrated the Court as a catalyst of EU regulatory reform. The Court has been described as an ‘information catalyst’ promoting procedures which generate deliberation among all relevant experts thereby ensuring that the science behind regulatory decisions meets the standards of independence, excellence, and transparency.⁴

In this paper, we re-evaluate the CJEU’s record in holding EU regulatory science to account over the last 20 years. We study the Court’s jurisprudence through a new, socio-legal lens by analysing the interactions between judicial review and the epistemic power of economic actors (ie the regulated industry)⁵ in EU risk regulation. We start with two important observations. Firstly, the CJEU modern approach to reviewing science-based regulation has been almost exclusively a response to litigation brought and framed by the regulated industry. The intensified judicial review, which resulted from that litigation, has reduced the EU administration’s discretionary space regarding factual findings and assessments and has mostly benefited economic actors who seek to challenge EU regulation.⁶ Secondly, the regulated industry plays a crucial role in the construction of EU regulatory science. Highly resourceful economic actors provide the scientific information on which decisions are based, comment on the decision-making process and litigate against decisions which in their view have not duly considered their information inputs. While the second observation has triggered debates about the risk of epistemic capture, the first observation has been largely neglected in debates about the role of the Court in risk regulation. In a first analysis of this kind, this paper examines the ways in which the CJEU doctrine on legal standing and procedural review of discretion is both shaped by the economic actors’ institutional position as information providers and further entrenches that position. In other words, we are interested in the institutional effects of judicial review on the epistemic power of economic actors, understood as the relative influence they can exert on what EU regulators believe, think, or know about health and environmental risks. This is an important question for EU expert accountability more generally. A lot has been written about the need for inclusive deliberation, epistemic diversity, and broad public scrutiny as ways to ensure non-biased, independent, and excellent regulatory science.⁷ EU regulation has been reformed over the years to provide many channels of public participation including at the stage of scientific assessments. Yet to what extent EU judicial review contributes to the participatory nature and broad public scrutiny of EU regulatory science remains an open question.

In recent years, there have been growing concerns over undue corporate influence on and epistemic capture of EU risk regulation. Numerous controversies, such as cases of conflict of interests in European agencies,⁸ the EU glyphosate re-authorisation amidst concerns over industry manipulated studies,⁹ the authorisation of endocrine disruptors,¹⁰ and the drafting of the EU bee guidance,¹¹ just to name a few examples, have reignited the debate about the adequacy of existing rules and practices regarding the composition and independence of expert bodies, the integrity of authorisation procedures, as well as about the role of economic actors in EU regulation. These controversies have triggered responses by the EU legislature – for example, a special parliamentary inquiry on the authorisation of glyphosate and, as a direct response to that, a recent reform of the EU General Food Law, which strengthens the rules on the transparency and independence of risk assessments.¹² The EU Ombudsman has also intervened in important ways.¹³ Debates about regulatory capture, however, tend not to engage with the role of courts in holding experts to account.¹⁴ Neither have debates on courts in risk regulation directly engaged with the issue of capture.

Our paper closes this gap. We argue that the challenge of ensuring a high-quality and publicly interested regulatory science requires a comprehensive view – one that bridges disciplinary debates and juxtaposes doctrinal developments with interdisciplinary insights about regulation. To achieve this task, we follow a hybrid methodological approach, which combines legal doctrinal analysis of over a decade of CJEU case law with a comparative institutional analysis.¹⁵ The latter analyses judicial review from the perspective of institutional choice, ie from the perspective of the question of who decides and what the relative merits of ‘always imperfect alternative decision-making processes’¹⁶ (eg judicial review v. the administrative process) are. We adopt an epistemologically¹⁷ informed variant of comparative institutional analysis, ie we focus on a particular problem among a wider set of concerns around the modern administrative state, namely the ability of economic actors to participate in and to influence the construction of scientific knowledge underpinning EU risk regulation at the expense of other actors.¹⁸ We then juxtapose this with the participation and epistemic influence of economic actors in EU judicial review of risk regulation. Overall, we seek to understand the institutional effects of certain doctrinal choices on the epistemic power of economic actors in EU risk regulation. This hybrid approach allows us to offer both an internal (doctrinal) and external (interdisciplinary) critique of the CJEU approach, and of its ultimate contribution to EU expert accountability.

When addressing these questions, we also investigate whether and to what extent the Court can be expected to address epistemic power imbalances in the production of EU regulatory science. The ways in which the epistemic foundations of EU risk regulation come together are complex and so are the ways in which judicial review interacts with these processes. Law is always only one part of broader institutional realities, and it follows its own doctrinal logic. The reason why the epistemic power of economic actors has not yet been addressed in discussions about judicial review of EU risk regulation is probably because this is assumed to be a question of empirical reality that falls outside the formal legal-normative realm. We aim to show however that at the very least doctrinal approaches can be complicit in either ignoring epistemic power imbalances in risk regulation or in failing to address them. If so, they are ill-adjusted to the reality of modern risk regulation and its challenges. As we show, judicial understandings of who should have access to the Court (legal standing) and to the administrative process (procedural rights and principles) as well as understandings of administrative discretion are all but neutral. They have consequences for the balance between private and public interests in EU regulation and for the quality of its scientific basis. In our view, any normative engagement with the case law must consider these consequences.

We develop our argument in four steps. We first set out the context for our case law analysis by explaining the ways in which the regulated industry exercises epistemic power over the EU regulatory process, and how the latter involves economic actors as key information providers (Section 2). We then explore how EU judicial review contributes to strengthening the role of economic actors as information providers thereby reinforcing their epistemic power. We focus on the Court’s approach, first, to the standing of private applicants (Section 3) and second, to the review of fact and discretion in litigation involving challenges to EU administrative rulemaking (Section 4). We analyse risk regulation case law between 2010 and 2020 to show the evolution of the Court’s post-Pfizer approach and of the three administrative principles which constitute the hallmark of the Court’s procedural review of regulatory science, ie the duty of care, the duty to state reasons, and the right to be heard. We then evaluate our findings combining doctrinal critique with an interdisciplinary perspective on expert accountability (Sections 5 and 6). We argue that, in contrast to previous expectations, the Courts’ current doctrinal understanding of procedural review fails to catalyse inclusive deliberation and, therefore, procedures able to secure independent and excellent scientific assessments. Coupled with an intensified review of technical discretion, the post-Pfizer modern approach is problematic, because it entrenches economic actors’ epistemic power at the expense of other relevant actors. We sketch out a normative vision for EU judicial review of risk regulation that, while being doctrinally coherent, would contribute to expert accountability in a more meaningful way.

2. The epistemic power of economic actors in EU risk regulation

In the EU, the qualitative and quantitative bulk of risk regulation takes place in the implementation phase of the policy cycle. Here, the Commission, assisted by a network of EU regulatory agencies, specifies the technical aspects of legislation.¹⁹ Despite the appearances, the implementation phase is not merely about technicalities: EU legislation is often formulated in an open-ended way,²⁰ and the definition of substantive, often contentious issues, is frequently allocated to non-legislative acts. Recurrent instances of politicisation, such as with pesticides and authorisations of genetically modified organisms (GMO), show that despite implementation’s lower political salience, stakes remain high.²¹

Economic actors play an important role in this context. As applicants, they apply for the authorisation of products, such as GMOs, chemicals, and pharmaceuticals, which may cause harm to consumers or the environment. As lobbyists, they seek, together with other interest groups, such as civil society organisations (CSOs), to influence EU legislation and its implementation.²² In both cases, economic actors act as information providers vis-à-vis EU regulators, by deploying the technical knowledge they generate with their economic activities (ie developing and manufacturing their products).²³ Such technical knowledge is a valuable asset, as risk regulation largely hinges on technical assessments. EU institutions, and in particular the Commission and European agencies, are notoriously understaffed²⁴ and therefore dependent on the technical information provided by external actors, ie private and public interest groups and economic actors.²⁵ Regulators’ informational needs are indeed one of the reasons for the involvement of economic actors in regulation.²⁶

By deploying their technical knowledge in a contested regulatory space, economic actors exert a form of epistemic power. Epistemic power is the ability to influence what others believe, think, or know and to enable and disable others from exerting epistemic influence.²⁷ In EU risk regulation, economic actors contribute to shaping both the regulators’ information base, by feeding information into the regulatory process, and their regulatory epistemology,²⁸ by promoting new regulatory paradigms, such as evidence-based regulation or the innovation principle.²⁹ In this way, they contribute to shaping the very way in which institutional actors think about risks and how to regulate them.

Economic actors exert epistemic power both when lobbying EU institutions and when acting as applicants for products authorisations. As lobbyists, economic actors have easy – albeit not exclusive – access to EU institutions. Empirical findings on lobbying in the implementation phase show that economic actors enjoy an advantage in terms of access to regulators vis-à-vis public interest groups.³⁰ Institutional efforts to enhance the latter’s access to EU decision-making are gradually creating a more level playing field,³¹ but public interest groups are still outnumbered by economic actors. The latter’s advantage has been found to be substantial regarding executive actors, in particular the European Commission and EU agencies, which dominate the implementation phase.³²

Generally, access to decision-making does not automatically translate into actual influence on policy outcomes.³³ Research on the legislative phase, for example, suggests that, notwithstanding economic actors’ advantages in terms of access, there is a relative balance in terms of the influence exerted by both economic and non-economic actors.³⁴ High political salience and the European Parliament (EP)’s involvement play to the advantage of public interest groups, mitigating access asymmetries.³⁵ In the implementation phase, however, political salience is low, the EP’s involvement marginal and policy-relevant technical information scarce. Economic actors’ chances to translate their access advantage into actual influence therefore increase.³⁶ There is ‘strong evidence to suggest that business groups attempt to mitigate any losses they might have incurred during the legislative phase’ through regulatory lobbying in the implementation phase.³⁷ In addition to regulators’ need for technical information, low political salience opens opportunities for ‘loophole lobbying’³⁸ with economic actors maximising the impact of the technical information they provide to the Commission and EU agencies.

Besides lobbying, regulatory design plays a role in entrenching economic actors’ epistemic power. Authorisation and registration procedures invite – or even require – economic actors to feed information into the regulatory process. Such procedures concern the allocation of health and environmental risks in society and often result in acts of general application: when a substance is approved to be included in a positive list, for example, it can be sold, bought, and used in the manufacture of products throughout the EU. At the same time, risk-regulatory procedures are individualised, meaning that they are set in motion by an individual producer (or by a group of producers) applying for the registration or marketing authorisation of a specific substance or product. This special character of risk-regulatory acts poses challenges in terms of the classical distinction in administrative law between individual and general determinations,³⁹ an issue to which we will return.

As applicants, economic actors are expected to include in the application dossier the scientific studies demonstrating their products’ compliance with the relevant regulatory requirements.⁴⁰ That is because in EU risk regulation, the burden of proof as to the safety of products is allocated to applicants. As a result, scientific assessments rely widely on the evidence generated, sponsored, and presented by the regulated industry.⁴¹ That is not to say that industry’s epistemic power is exclusive or unfettered. Publicly available scientific literature is also included in the application dossiers⁴² and legal safeguards are in place to ensure the scientific quality and independence of the studies relied on.⁴³ However, concerns over the independence of industry studies remain, as shown by the glyphosate controversy, and legal safeguards have been found to be insufficient, either by design or because of their inconsistent application.⁴⁴

Economic actors’ epistemic power is deployed throughout the authorisation procedure. They maintain frequent contacts with the regulators, regularly exchange correspondence with and attend the expert meetings of EU and Member States’ agencies.⁴⁵ Most regulatory frameworks have sought to actively widen the range of actors involved, by providing opportunities for public participation (eg through public consultations).⁴⁶ However, economic actors’ role as applicants reinforces their epistemic agency as compared to other actors. This poses accountability and legitimacy issues as their involvement as information providers might result in biased regulatory outcomes.⁴⁷ We refer to this as the risk of a ‘minoritarian bias’, ie the ‘over-representation of concentrated interests to the detriment of majorities’.⁴⁸

The above discussion does not prove the actual existence of a minoritarian bias in EU risk regulation. Yet, it confirms that such risk is present. Economic actors exert epistemic power over EU institutions on technical matters, by feeding information into the regulatory process, and by strategically mobilising their epistemic and other resources to push their (de)regulatory agendas. In the implementation phase, this occurs both de facto, through lobbying, and de jure, through legal anchoring of their position as information providers. This is the context, in which industry challenges to EU risk regulation take place. In the following sections, we consider how, in the adjudicative phase, economic actors’ epistemic power meets the CJEU’s understanding of the administrative process, starting with its doctrine on legal standing.

3. Access to the Court – who gets to challenge EU regulatory science?

The two most salient features of EU risk regulation litigation are its prominent focus on regulatory acts and its being quantitatively dominated by economic actors. Between 2010 and 2020, European Courts (Court of Justice and General Court) decided on the merits of 102 validity challenges (actions for annulment and preliminary rulings on validity) to risk regulation measures.⁴⁹ The vast majority of these judgements (89 out of 102) concerns actions for annulment against non-legislative (ie implementing and delegated) measures,⁵⁰ most of which are administrative acts of general application, with a minority of individual acts.⁵¹ Of these, 68 were initiated by economic actors, 7 by Member States, and 14 by public interest litigants. The divide is even starker when excluding access to documents litigation⁵² and actions brought by Member States,⁵³ with economic actors initiating 60 challenges against 4 being brought by non-privileged public interest applicants.⁵⁴ Economic actors’ dominance is partly a matter of resources. Compared to public interest groups, economic actors are financially better equipped to secure access to legal expertise⁵⁵ and engage in costly and frequently lengthy legal proceedings.⁵⁶ In addition, litigation is an increasingly appealing strategy for economic actors.⁵⁷ The judicial declaration of a measure’s illegality results in a revocation of an ‘unfavourable’ rule altogether. Even more importantly, it may affect future regulatory outcomes by producing an anticipatory chilling effect,⁵⁸ with the Commission being more attentive to concentrated interests’ arguments than to those of dispersed interests and thus reinforcing the former’s epistemic power.

Such a polarised litigation pattern, however, is not a matter of resources and strategy alone. The Court’s interpretation of legal standing reinforces the industry’s position as litigant and, in doing so, enables the propagation of its epistemic influence from the administrative into the adjudicative phase. In this section, we analyse how the position of the regulated industry as applicants in authorisation procedures also translates in their privileged access to judicial review as compared to other actors who are affected by and might want to contest the science behind EU regulation.

The Court’s strict interpretation of standing requirements is subject to a long-standing scholarly debate, and we must limit ourselves to the main points relevant for this paper’s argument. The bone of contention is the interpretation of Article 263(4) TFEU, which happens to be the main procedural gateway to challenge EU risk regulation measures.⁵⁹ According to this provision, the standing of plaintiffs other than the addressee of an act is subject to the double requirement of direct and individual concern. Since its ruling in Plaumann in 1963, the Court has held that ‘[i]n order for a measure to be of individual concern to the persons to whom it applies, it must affect their legal position because of a factual situation which differentiates them from all other persons and distinguishes them individually in the same way as a person to whom it is addressed’.⁶⁰ Direct concern, on the other hand, has been interpreted as meaning that the plaintiff’s legal situation must be affected as a direct consequence of the contested act, without the intermediation of additional rules.⁶¹The strict interpretation of individual concern significantly curtails the possibility for persons other than the (quasi-)addressee to bring actions for annulment. This is especially problematic for public interest groups, such as CSOs or patients’ associations.⁶² In Granosalus, an association of wheat producers and consumers seeking to challenge glyphosate’s authorisation was found not to be individually concerned, because ‘the contested act affects the applicant’s members by reason of their objective status as consumers, citizens of the EU or wheat producers in the same way as any other consumer, citizen of the EU or wheat producer who is actually or potentially in the same situation.’⁶³ Associations and public interest groups are therefore structurally incompatible with the individualised approach followed by the Court,⁶⁴ no matter how strongly they are affected by the decision at hand.

The Lisbon Treaty has partly relaxed the rules on locus standi for regulatory acts, to the effect that private applicants can challenge ‘a regulatory act which is of direct concern to them and does not entail implementing measures’.⁶⁵ The Court interprets the term ‘regulatory act’ as referring to non-legislative acts⁶⁶ of general application.⁶⁷ However, this partial relaxation has not brought about significant change in the ability of interest groups to challenge regulatory acts. The direct concern requirement remains an obstacle. It requires that the act must directly affect the plaintiff’s legal situation and leave no discretion as to its addressees, who are entrusted with the task of implementing it, such implementation being purely automatic.⁶⁸ Risk regulatory acts, such as positive lists of authorised substances and products, often require further implementation at the national level or are part of two-stage procedures, such as in the case of pesticides. In Pesticide Action Network Europe v Commission, a group of environmental CSOs challenged the inclusion of sulfoxaflor in the list of authorised active substances under the Pesticides regulation and was denied standing due to the lack of direct concern. The Court found that the contested act, a Commission implementing regulation⁶⁹ and an act of general application, only affects the legal situation of the Member States (enabled to authorise the placing on the market of products containing the active substance) and that of potential applicants for such authorisations. According to the Court, the act had ‘neither the purpose nor the consequence’ of granting rights or imposing obligations on other subjects.⁷⁰

Notwithstanding the Lisbon amendment, the Court invariably draws the circle of both individually and directly concerned applicants very narrowly: as of 2021, no public interest group has been considered directly concerned by the risk regulation measure it sought to challenge via Article 263(4).⁷¹ Despite widespread criticism from academia,⁷² international organisations⁷³ and the Court’s Advocate Generals,⁷⁴ the Court has not been willing to change its interpretative stance.

The Court’s approach reinforces the regulated industry’s position as main information provider. Economic actors exert epistemic power by feeding information into the regulatory process and they are the only actors, according to the Court, who are entitled to ask for a review of whether such information was duly considered by the administration. In the next section we look at how the Court reviews administrative discretion in risk regulation. We show that a similarly narrow and individualised understanding of administrative procedures characterises its doctrinal approach to the review of regulatory science. Like with locus standi, this ultimately privileges the regulated industry and entrenches its epistemic power in both the regulatory process and its judicial review.

4. Judicial review of EU regulatory science – technical discretion and the function of administrative process guarantees

A. The CJEU’s modern approach – the synergy between substantive and procedural review

The CJEU has played an important role in shaping the EU’s approach to risk regulation,⁷⁵ particularly in the aftermath of the bovine spongiform encephalopathy (BSE) crisis and the ensuing overhaul of EU food and product safety legislation.⁷⁶ The case law developed against the background of growing powers of the EU administrative state characterised by the transfer of more legislative as well as direct administrative powers to the EU level,⁷⁷ the growth of EU agencies,⁷⁸ and the increasing techno-scientific complexity of administrative decision-making on health, safety, and the environment. In response, the Court has gradually defined its modern approach to the review of discretion in risk regulation, thereby developing principles that ought to govern the intricate interplay between science, politics, and law in this field.

EU risk regulation case law, therefore, is often understood as a response to a maturing and expanding EU administrative state. It is often neglected, however, that this very understanding has been reinforced by the way in which economic actors have framed their litigation before the Court. As discussed above, industry challenges against EU administrative rulemaking dominate the litigation. Such challenges are typically framed along the lines of a classic (neo-)liberal conception of the administrative state, which emphasises the dangers of an administration unbound,⁷⁹ ie of too much intrusion into the private sphere and freedom of citizens. Industry challenges tend to focus on concerns over arbitrary or biased use of administrative power, allegedly not grounded in ‘sound science,’ but in political opportunism or faulty public perceptions of risk.⁸⁰ Wide discretion and the precautionary principle, both deeply interrelated, are in this conception the symptoms of an ailing regulatory system, and the antidote presented in response is evidence-based regulation.⁸¹ Control of the exercise of administrative discretion and the need for ‘evidence-based’ decisions are seen as crucial in avoiding administrative overreach into individual rights of citizens, or in this context, the industry’s economic rights.

As a result of industry litigation, the Court’s case law over the last twenty years has been focused on addressing such concerns and showed to be responsive to them. EU risk regulation case law is characterised by an emphasis on scientific evidence as a check on the exercise of administrative discretion as well as the willingness of the Court to review such exercise more intensely. Overall, two main features define the Court’s modern approach.⁸² The first is an intensified substantive review of discretion. Looking at the evolution of the case law, there has been a significant shift from a highly deferential, light-touch approach,⁸³ to a more intense scrutiny of the substance of risk regulatory measures. This applies particularly to the review of facts and of assessments of facts, and hence of the scientific basis underlying decisions.⁸⁴ This shift, perhaps surprisingly, did not occur because of the Court explicitly changing the applicable standard of review, but in a subtler way through the Court’s re-interpretation of the applicable standard.⁸⁵

In both early and contemporary case law, the Court has continuously applied the same ‘manifest error of assessment’ test to the review of both fact and discretion. The mantra repeated in much of the case law is that in cases of high techno-scientific complexity the administration enjoys a broad discretion, which the Court is bound to respect. Hence, the Court should not substitute its judgement for that of the EU administration on questions of fact and discretion and only carry out a limited review. Instead, the party challenging the rule must show that the latter is ‘vitiated by a manifest error or misuse of powers’ or that the administrative body has ‘manifestly exceeded the limits of its discretion’.⁸⁶ However, the meaning of this test has changed over time. In early cases, such as Fedesa, the Court would not engage with the science underpinning the contested measures, typically rejecting the allegation of manifest error in a brief manner unless some flagrant error could be shown on the face of the decision.⁸⁷ This changed around the turn of the millennium, when cases such as Pfizer, ⁸⁸ Alpharma,⁸⁹ and Artegodan ⁹⁰ signalled the opening of a new chapter in the Court’s case law. From then on, the Court typically ascertains whether there is a manifest error by engaging in a more searching review of both the scientific evidence and the administrative reasoning.⁹¹

The second feature of the CJEU’s modern approach is its procedural nature. The intensification of substantive review via the manifest error of assessment test went hand in hand with – or rather, was achieved through – an emphasis on procedural standards flowing from the right to a good administration, most notably the duty of a careful and impartial examination, the duty to give reasons, and the right to be heard.⁹² The Court’s reliance on procedural guarantees emerged as a form of compensation for its limited review of the merits in cases entailing complex technical assessments. This approach was first articulated in Technische Universität München,⁹³ where the Court stated that:

since an administrative procedure entailing complex technical evaluations is involved, the Commission must have a power of appraisal in order to be able to fulfil its tasks. However, where the Community institutions have such a power of appraisal, respect for the rights guaranteed by the Community legal order in administrative procedures is of even more fundamental importance. Those guarantees include in particular the duty of the competent institution to examine carefully and impartially all the relevant aspects of the individual case, the right of the person concerned to make his views known and to have an adequately reasoned decision.⁹⁴

Procedural guarantees acquired a particular meaning in the field of risk regulation, as the Court clarified in Pfizer. Here, it established that:

a scientific risk assessment carried out as thoroughly as possible on the basis of scientific advice founded on the principles of excellence, transparency and independence is an important procedural guarantee whose purpose is to ensure the scientific objectivity of the measures adopted and preclude arbitrary measures.⁹⁵

Commentators have highlighted that, under the label of procedural review, the Court started using substantive and procedural standards in a synergetic way.⁹⁶ In other words, procedural standards invite the Court to engage with risk assessments, which in turn facilitates substantive scrutiny, leading to an overall intensification of judicial control over EU regulatory science.

The modern approach has been consolidated in the subsequent case law. The intensified scrutiny under the manifest error of assessment test⁹⁷ and procedural review⁹⁸ are now constant features of EU risk regulation jurisprudence and have been widely discussed among legal scholars. The modern approach has been dubbed as substantially deferential, but procedurally activist,⁹⁹ because it would allow the Court to respect the administration’s discretion, hold the decision-makers to account for the way in which they collected and used scientific information, as well as review the adequacy of administrative reasoning based on that evidence, all at the same time. In this way, the Court’s case law has often been seen as achieving a sort of Goldilocks balance between the Court’s task of ensuring judicial protection, on the one hand, and its duty, flowing from the principle of institutional balance, to respect the administration’s prerogatives as decision-maker, on the other hand.¹⁰⁰

On some accounts, the Court’s approach in cases such as Pfizer, ¹⁰¹ Alpharma,¹⁰² and Artegodan, ¹⁰³ is seen as having a more far-reaching catalytic effect on the legitimacy of the EU administrative process. In this view, the Court’s case law is seen as part of a broader governance trend whereby judicial review fulfils a quasi-regulatory function by promoting the adherence to deliberation-enhancing procedural standards, such as participation, inclusion, and transparency.¹⁰⁴ With regard to regulatory science, the Court acts as an ‘information catalyst’ seeking to ensure, through the use of the principles of care and reason-giving, that the evidence underpinning risk decisions is of high epistemic quality.¹⁰⁵ In Corkin’s words, the Court is ‘the “catalyst” that enhances [decision makers’] capacity to regulate legitimately by articulating norms that prompt them to involve affected parties, regulate from an adequate informational base and do so in a principled (deliberative), accountable, and transparent manner’.¹⁰⁶

While the Court’s case law does indeed catalyse procedural principles, the extent to which the use of such principles serves to involve all affected parties and to promote inclusive deliberation in EU risk regulation is doubtful. Studies since Pfizer have mostly attempted to make sense of the case law or to evaluate it either from the deliberative governance point of view or from the judicial protection of the applicant’s rights point of view or a mixture of both. While we do not negate the importance of both deliberation in the administrative process and judicial protection of rights, we are concerned about the narrow circle of beneficiaries drawn by the Court’s doctrinal approach and the potential institutional consequences for the epistemic power economic actors are able to exercise over both the regulatory and judicial process. In fact, the role of the industry in both shaping the Court’s approach and benefiting from it has remained under the radar so far. As we show in the following, the way procedural grounds of review, such as the duty of care, the duty to state reasons, and the right to be heard, are understood and utilised in the case law makes them an effective mechanism in the hands of the regulated industry to challenge EU agency science and therefore public regulation, while no comparable mechanisms are available to other actors, most notably civil society. The consolidated litigation strategy of the regulated industry pushes the Court to intensify its review of the factual basis underpinning the administration’s decision. Vanda, a case concerning the authorisation of the schizophrenia treatment Fanaptum, is emblematic of how industry applicants exploit the synergy between a more searching scrutiny and procedural review:

the applicant submits that the Commission is seeking, by putting forward an excessively restrictive interpretation of the extent of judicial review, to adopt a ‘strategy of obfuscation’ intended to convince the Court not to examine the merits of the pleas and to prevent it from addressing several fundamental questions concerning the compatibility of the contested decision with EU law. The applicant submits that it is apparent from the case-law that the EU judicature is in a position to assess the legality of the scientific evaluation made by the CHMP [Committee for Medicinal Products for Human Use] […] and, where appropriate, to determine whether the unlawfulness of that assessment constitutes a breach of an essential procedural requirement rendering the Commission’s decision unlawful.¹⁰⁷

In the following, we look more closely into how the Court understands the purpose and scope of the process values it deploys in responding to industry challenges. We consider the actions for annulment, excluding cases on access to documents, brought by non-privileged applicants between 2010 and 2020 against administrative rulemaking and decided on the merits. Here, the most frequently invoked grounds of review are (manifest) errors of assessment (33 times) and breach of the principle of proportionality (27 times). Other frequently invoked grounds are the duty to state reasons (14 times); the rights of the defence (including the right to be heard and the right to a fair hearing) (14 times); the principle of good or sound administration (10 times); and, although more rarely, the precautionary principle (5 times). Being concerned with economic actors’ epistemic power, our focus is on challenges to risk assessments (ie the epistemic foundation of risk regulation measures), occurring mainly through complaints about manifest errors of assessment and the invocation of the procedural principles of duty of care, reason giving, and the right to be heard.¹⁰⁸ We look at how these principles have been understood and utilised in the case law between 2010–2020 and how that affects the epistemic power of the regulated industry.

B. Case law 2010–2020: To whom does the EU administration owe care?

In the EU legal order, the duty of care is ‘the duty of the administration to impartially and carefully collect and examine the information needed for its decision-making’.¹⁰⁹ In risk regulation it has been interpreted by the Court as entailing the administration’s duty to carry out a thorough risk assessment based on the principles of excellence, independence, and transparency.¹¹⁰

In the context of administrative rulemaking, the duty of care is an objective procedural requirement which ‘contributes to a factually accurate decision and thereby to achieving the public interests laid down by its legislative mandate’.¹¹¹ In Arizona Chemical, a case concerning the exclusion of a chemical substance from the positive list of authorised substances, the Court explicitly differentiated between the role of the duty of care in procedures resulting in the adoption of an administrative act of individual application as opposed to acts of general application. In the former, following the T.U. München approach,¹¹² the duty of care operates as a procedural guarantee individuals can invoke against discretionary acts which adversely affect their legal sphere; in the latter, the duty of care is ‘essentially an objective procedural guarantee arising from an absolute and unconditional obligation on the Community institution relating to the drafting of an act of general application and not the exercise of any individual right’.¹¹³ The differentiation in Arizona Chemical between the objective and individual dimensions of the duty of care depending on the act to which it is applied remains rare. As we discuss in the following, the Court has settled for the T.U. München approach to the duty of care also when reviewing acts of administrative rulemaking, emphasising its protective and compensatory dimension to the benefit of the individual applicant, rather than its objective and public interest nature.

As mentioned in the previous section, the Court’s review based on the duty of care is enmeshed with the substantive review of administrative discretion. Duty of care is rarely invoked explicitly as a self-standing ground of review but is rather mediated through the manifest error of assessment test. This has been the case since Pfizer:

The Community judicature is not entitled to substitute its assessment of the facts for that of the Community institutions, on which the Treaty confers sole responsibility for that duty. Instead, it must confine itself to ascertaining whether the exercise by the institutions of their discretion in that regard is vitiated by a manifest error or a misuse of powers or whether the institutions clearly exceeded the bounds of their discretion. […] However […] in such circumstances, the guarantees conferred by the Community legal order in administrative proceedings are of even more fundamental importance. Those guarantees include, in particular, the duty of the competent institution to examine carefully and impartially all the relevant aspects of the individual case.¹¹⁴

When assessing whether a manifest error of assessment has been committed by the EU administration, the Court does so by checking whether all the relevant evidence has been carefully and impartially considered ie whether the duty of care has been fulfilled.¹¹⁵ Notwithstanding the Court’s stated limited review, the procedural scrutiny carried out under the manifest error of assessment test has enabled the Court to go further into the review of regulatory science. In Pfizer, it was used by the Court to thoroughly (230 paragraphs, 71 pages) engage with scientific arguments of the parties in order to review the substantive rationality (ie their non-arbitrary nature) of the administrative rule in question. It scrutinised the presence of factual errors, the adequacy of the scientific data, as well as the reasoning of the Commission when assessing the scientific evidence and drawing conclusions from it.¹¹⁶

Since Pfizer, the Court’s framing of its review of complex technical assessments has become more sophisticated and demanding. Over the last decade, the Court has started qualifying the manifest error of assessment test by formulating a plausibility requirement. Since the early 2010s, it has increasingly required that

in order to establish that the Commission committed a manifest error of assessment in assessing complex facts such as to justify the annulment of a decision which is contested, the evidence adduced by the applicant must be sufficient to make the factual assessments used in the decision implausible.¹¹⁷

Arguably, plausibility requires a subtler engagement with the substance of the case, compared to a ‘simple’ manifest error of assessment test. The Court’s duty to control whether the relevant aspects of the individual case have been carefully and impartially examined has developed into a more refined scrutiny of:

• whether the evidence relied on is factually accurate, reliable, and consistent,

• whether that evidence contains all the information which must be considered to assess a complex situation, and

• whether it can substantiate the conclusions drawn from it.¹¹⁸

While intensifying the scrutiny in this way, the Court continues to emphasise the protective individual dimension of the principle of duty of care also when reviewing acts of administrative rulemaking. The Court invariably reiterates the reasoning according to which wide discretion entails limited review, limited review entails enhanced procedural scrutiny, and procedural review entails ‘the obligation to examine carefully and impartially all the relevant elements of the individual case’ to ensure the scientific objectivity and lack of arbitrariness of the measures adopted.¹¹⁹ Rather than operating as an objective procedural guarantee, aimed at protecting the public interest in good administration, the duty of care is seen as protecting the applicant against arbitrary exercises of administrative discretion.

Let us now consider, first, the type of scenarios that prompt the Court’s procedural review based on the duty of care and, second, how the Court carries it out in practice. Economic actors challenging the scientific basis of regulatory measures lament two main forms of manifest errors of assessment: the risk assessor (ie the relevant agency) did not consider all the relevant evidence; the assessment methods adopted by the risk assessor are not adequate. In the following, we analyse the Court’s response to these challenges.

B.1 Challenges to the completeness of the information basis

The first type of manifest errors of assessment invoked by economic actors is the incompleteness of the information underpinning the contested measure. According to litigants, such incompleteness indicates a manifestly erroneous or implausible assessment where the agency has not carefully considered all the relevant information, typically studies submitted by the applicant in the application, or has failed to carefully consider the comments submitted by the applicant during public consultations or other exchanges with the agency. Failure to consider such evidence, so the argument goes, affects the assessment’s correctness, in so far as, if duly considered, such information would have led to a different regulatory outcome. This type of challenges results in the Court engaging in a searching review of the administration’s determinations. While in most cases the Court acknowledges the administration’s wide discretion,¹²⁰ it reaches this conclusion after having dissected the scientific underpinnings of the contested decision.

Such dissection can take the form of a box-ticking exercise, aimed at verifying whether the allegedly neglected studies formally feature in the contested risk assessment,¹²¹ or engage directly with its substance. Bayer ¹²² is a good example of the latter. Here, several pesticides producers challenged the Commission’s decision to prohibit the sale and use of seeds treated with plant protection products containing three active substances deemed to be lethal to bees. The applicants lamented the lack of consideration of specific studies and monitoring data included in their applications. When reviewing EFSA’s opinion, on which the contested decision was based, the Court assessed whether and how the agency considered the applicant’s data. It found that the European Food Safety Authority (EFSA) had taken such data into account but that, nevertheless, such data could not ‘have a decisive influence on the outcome of the risk assessment and, in particular, cannot establish with sufficient certainty the safety of the substances covered’.¹²³ Hence, there was no manifest error of assessment. This shows that the Court holds regulatory science to account by scrutinising the agency’s use of evidence but does so only with respect to the scientific studies invoked by the industry applicant.

Paradoxically, the Court’s exclusive approach to the duty of care applies also to the evidence submitted in the context of public consultations, which are specifically aimed at broadening the range of information considered in the regulatory process. Several risk regulation regimes provide for public consultations and, in general, economic as well as other actors can submit information to risk assessors. However, only the industry applicants can challenge unfavourable regulatory outcomes on the ground that such information has not been duly considered.

A typical scenario is that in ICL-IP.¹²⁴ The case concerned the classification of a chemical compound, nPB (1-bromopropane), as a substance of very high concern according to Article 57(c) REACH. Among other grievances, the applicants lamented that the Commission had breached the principle of sound administration and committed a manifest error of assessment by ‘relying exclusively on the data concerning the volumes contained in the registration file and not on the data submitted in the context of the public consultation and, subsequently, during the procedure leading to the adoption of the contested regulation’.¹²⁵ The applicants had actively interacted with the European Chemicals Agency (ECHA) and the Commission by submitting comments and letters supporting the case for nPB’s non-inclusion as a substance of very high concern. In so doing, they had submitted new information, some of which contradicted that originally included in the registration dossier. The Court considered the issue extensively (in a total of 58 paragraphs). First, it reviewed ECHA’s responses to the applicants’ public consultation comments to establish whether it had considered that additional information.¹²⁶ It found that the agency had replied to the comments and established that they did not call into question its conclusions.¹²⁷ Second, the Court considered whether the Commission had committed a manifest error of assessment by not using the additional information submitted in the context of the public consultation. Here, the Court concluded that the EU administration could not be criticised for having relied primarily on the information included in the registration dossier submitted by the applicant, instead of that submitted, again by the applicant, during the public consultation. It should be noted that references to information submitted by other actors are absent from the Court’s reasoning. Even when embedded in participatory mechanisms such as public consultations, the Court uses the principle of duty of care only to ascertain that the decision-maker considered information submitted by the applicant. While this might not be surprising, as the Court is reacting to the applicants’ requests, it obscures the fact that the Court could, by acting ex officio, use the duty of care to scrutinise the decision’s epistemic basis in a more comprehensive fashion, reaching beyond the applicant’s requests and verifying instead whether all the information submitted during the public consultation has been duly considered.¹²⁸ As of today, however, the Court’s focus remains narrow: in the Court’s words, ‘it is true that the public consultation provided for in Article 58(4) of Regulation No 1907/2006 enables all interested parties to provide relevant data. It therefore enables the registrants to produce additional information concerning the uses to be exempted from the authorisation requirement.’¹²⁹

The judgements dealing with allegations of incompleteness confirm that despite the Court’s expressed respect for the administration’s broad discretion concerning factual assessments, a relatively intense scrutiny of such assessments takes place under the principle of duty of care. The latter is, moreover, interpreted as protecting the interests of the applicant rather than an objective procedural guarantee. In theory, careful and impartial consideration is owed to ‘all the relevant evidence’. In practice, the dissection of the science behind the contested measures is conducted exclusively on the grounds of a limited set of evidence, ie that submitted by economic actors as applicants in the administrative process and subsequently invoked by them as litigants in the adjudicative phase. Rather than promoting a more epistemically diverse information basis, the Court’s enhanced scrutiny is limited to double-checking whether the information submitted by the same narrow circle of actors has been duly considered.

B.2 Challenges to the choice of the assessment method

The cases considered above respond to legal challenges pivoted on the allegation of incomplete information. A second line of complaints based on the principle of duty of care challenges the scientific assessment methods chosen by the administration. Methodology cases are characterised by a dense intertwinement of scientific and legal considerations, whereby the Court reviews whether the chosen assessment method is allowed by the relevant legislation and correctly applied. They are often ambiguous and unpredictable in terms of standard of review,¹³⁰ but overall they confirm the Court’s exclusive reliance on the evidence invoked by the applicants.

Bilbaìna v Commission ¹³¹ epitomises the pitfalls of methodology review and is one of the few cases resulting in the annulment of the contested measure, in so far as the Commission erred in the choice of the assessment method. Several producers sought the annulment of the Commission’s classification of pitch, coal tar, high-temp (CTPHT), as a substance of very high concern under the Regulation on the classification, labelling, and packaging of substances and mixtures (CLP Regulation) due to its aquatic toxicity. They challenged the method applied to the assessment of the substance’s properties, which allegedly resulted in an unreasonably unfavourable assessment.¹³² The Court upheld the complaint, because the Commission had ‘failed to comply with its obligation to take into consideration all the relevant factors’. These included elements, such as ‘the proportion in which the 16 polycyclic aromatic hydrocarbon (PAH) constituents are present in CTPHT and their chemical effects,’ namely their low water solubility, which were suggested by the applicants, but not mentioned by the regulation.¹³³ As a result, the Court found the Commission’s assessment to be ‘non realistic’.¹³⁴ Bilbaìna v Commission clearly shows, that the Court’s modern approach can result in an intrusive review with the Court close to substituting its assessment to that of the Commission. It also shows that such intrusive scrutiny results from the litigants’ operationalisation of the duty of care, to which the Court responds.

‘Methods review’ cases invite the Court to engage with the quality of the risk assessment process. Instead of taking a comprehensive view of the decision-making process and its deliberative qualities, its scrutiny remains firmly anchored within the applicant-administration binomial. At most, the Court widens the circle of information providers slightly – and rarely – by referring to peer-reviewed scientific literature, but comments and studies submitted by third parties in the context of public consultations remain out of the Court’s sight. Therefore, regrettably, the Court’s engagement with the risk assessment process does not catalyse deliberation, at least not in its proper sense of including all those concerned.¹³⁵ The Court is satisfied with the fact that the relevant scientific committee, which often includes industry representatives, bases its opinion ‘on the results of a number of tests carried out using different methods, confirmed by information taken from specialist publications’.¹³⁶ There is no apparent concern for the process including a wider range of relevant voices. This is problematic, in light of the epistemic power exerted by economic actors in the regulatory phase and of the intensified review of risk regulation measures in the adjudicative phase. In other words, the Courts’ more stringent review mostly benefits those actors, ie the regulated industry, who already enjoy a dominant position in the regulatory process.

It is important to note that the judicial emphasis on process values is not only a reaction to industry-led litigation, but also a consequence of the Court’s attempt to compensate for the difficulty of substantive review in highly technical matters.¹³⁷ This results in a narrowing down of the function of principle of duty of care in procedures of general rulemaking, a point to which we will return at the end of this paper.

C. Case law 2010–2020: To whom does the EU administration owe justification?

The duty to state reasons, as enshrined in Article 296 TFEU, entails that ‘legal acts shall state the reasons on which they are based.’ The duty to state reasons has a threefold dimension: protective (giving ‘an opportunity to the parties to defend their rights’), control (allowing ‘the Court to exercise its supervisory functions’) and of objective procedural guarantee (giving ‘an opportunity to Member States and to all interested nationals of ascertaining the circumstances in which the Commission has applied the Treaty’).¹³⁸ In its objective dimension, the duty to state reasons provides ‘the basis for efficient and fair administrative decision-making by ensuring that officials perform their tasks free from outside interference and arrive at decisions in a rational way’.¹³⁹ More generally, reason giving fulfils a broader democratic function by allowing the public to understand and scrutinise the exercise of public power.¹⁴⁰ It applies to all EU legal acts, regardless of their individual or general application.

Like the duty of care, the duty to state reasons is invoked by litigants to support allegations of unsoundness, incompleteness, or inconsistency of the administration’s assessments. Rephrasing the same complaints in terms of reason-giving, economic actors argue that the administration has not sufficiently articulated the reasons for why it omitted their comments or studies,¹⁴¹ or for drawing certain conclusions rather than others from the evidence submitted.¹⁴² Thereby they again invite the Court to probe the scientific basis of contested measures.

While responding to similar complaints by the applicants, reason-giving review of administrative rulemaking is overall less intrusive than the plausibility review under the duty of care.¹⁴³ In terms of standard of review under the duty to state reasons, the Court consistently holds that the statement of reasons ‘must be appropriate to the act at issue and must disclose in a clear and unequivocal fashion the reasoning followed by the institution which adopted the measure in question’.¹⁴⁴ In the case of acts of general application, this is limited to explaining the general objectives the measure is intended to achieve and the general situation which led to its adoption, since ‘it would be excessive to require a specific statement of reasons for the various technical choices made.’¹⁴⁵ Hence, the administration’s statement of reasons does not need to address all the relevant facts and points of law, but only those ‘having decisive importance in the context of the decision’.¹⁴⁶ In general, the Court’s approach to reason giving has not so far resulted in a significant intensification of the justificatory threshold the administration is expected to meet. Furthermore, it signals an attempt to separate (substantive) challenges based on alleged errors of assessment and (procedural) challenges based on the duty to state reasons, the former targeting ‘the substantive legality of a decision’, the latter alleging ‘absence of reasons or inadequacy of the reasons stated’.¹⁴⁷ This general trend is however qualified by cases in which the Court engages in a detailed scrutiny of the reasons provided in scientific opinions: here, procedural and substantive scrutiny conflate, and reason-giving review largely replicates the manifest error of assessment test.¹⁴⁸

Like with duty of care, the Court stresses the protective individual nature of the reason-giving requirement also in procedures of general rulemaking. The Court conceives the purpose and beneficiaries of the principle of reason-giving in very narrow terms. In ICL-IP, it made clear that

The purpose of the obligation to state the reasons on which an act adversely affecting an individual is based, which is a corollary of the principle of respect for the rights of the defence, is (i) to provide the person concerned with sufficient information to make it possible to ascertain whether the act is well founded or whether it is vitiated by a defect which may permit its legality to be contested before the Courts of the European Union and (ii) to enable the latter to review the legality of that act.¹⁴⁹

The Court therefore focuses exclusively on the individual applicant, who should be given sufficient elements to understand and eventually challenge the measure affecting its individual sphere. The Court sees reason-giving and participation rights as closely related. The possibility to participate in the decision-making process has implications for the level of detail required of reason-giving, to the effect that ‘where the persons concerned are involved in the process by which a measure comes about, the requirement to state reasons may be circumscribed, since they acquire information through their involvement.’¹⁵⁰ Because of its exclusive focus on the applicant, the Court is satisfied with the latter’s actual participation in the administrative process, which then lowers the threshold for the exhaustiveness of reason giving. Conversely, the Court neglects the objective procedural nature of the principle of reason giving and its potential to open decision-making, if only ex post, to a broader range of actors.¹⁵¹ Reasons are due to those who already have access to the decision-making process, but not to those who are excluded from it. This is a missed opportunity in a field of high techno-scientific complexity, where judicial review could indirectly play the role of a translator, facilitating public understanding of the reasons underpinning complex measures.

D. Case law 2010–2020: Whose voice ought to be heard by the EU administration?

The rights of the defense (ie the right to be heard and the right to a fair hearing) are the third ground of review routinely invoked by litigants challenging EU risk regulation measures. Economic actors invoke violations of these rights in three main scenarios, which resemble closely those considered in relation to the duty of care and to the duty to state reasons. Litigants claim that 1) they have not been involved in the procedure in a way to would have enabled them to effectively defend their case; 2) they were not able to provide the necessary information; or 3) the relevant agency did not accurately consider the data submitted.¹⁵²

In the EU legal order, the right to be heard forms part of the right to a good administration as enshrined in Article 41 CFR. It entails ‘the right of every person to be heard before any individual measure which would affect him or her adversely is taken’. In the Court’s case law, the right to be heard applies, as a general principle, only to individual acts, and not to acts of general application.¹⁵³ In Vecco, the Court clarified that the procedure for the inclusion of substances in Annex XIV REACH ‘cannot be regarded as “proceedings initiated against” the applicants’ and neither ECHA nor the Commission are therefore required ‘to hear an individual who might be concerned by the disputed measure’.¹⁵⁴ In the case of individual acts, the right to be heard can be invoked by ‘the addressees of decisions which significantly affect their interests’.¹⁵⁵ Therefore, where authorisation decisions are addressed at an economic actor, such as with medicines, the latter can invoke the right to be heard before the Court. This, however, is not the case for other affected actors, even where those actors were consulted during the administrative procedure. In Olivieri, the applicant (a scientist) had participated in the procedure by submitting on her own initiative her views on the safety and efficacy of the medicinal product being authorised. The Court acknowledged that

None of the provisions of the applicable Community rules prohibits the Commission, prior to granting a marketing authorisation for a medicinal product, from following a procedure during which persons other than the applicant for marketing authorisation are able to submit their observations so as to enable it to fulfil its duty to check, in the interest of public health, that all the information relating to the scientific evaluation of the product in question, whether it be favourable or unfavourable to the product, has indeed been made available to it.¹⁵⁶

Yet this did not suffice to consider the applicant as a ‘formally concerned’ party in the administrative procedure, which resulted in the inadmissibility of her action.¹⁵⁷

Despite the right to be heard being confined to individual acts, the Court has applied it to acts of general application on several occasions. The condition for this, however, remains tied to the requirement that only persons who are directly and individually concerned by the act in question can benefit from the right to be heard. In this way, the Court’s approach to the right to be heard mirrors that to legal standing, and both must be considered together. Take, for example, Commission implementing decisions addressed to the Member States concerning the inclusion of active substances in so-called positive lists of authorised products.¹⁵⁸ These acts result from procedures initiated by economic actors, who, although not the addressees of the final decision (addressees being the Member States), are found to be directly and individually concerned due to their prior formal procedural involvement. Here, regulatory design meets the Court’s doctrine on standing, closing the circle between participation in the administrative and the adjudicative phase. The circularity is evident in SeproEurope, where the Court acknowledged that the applicant is not an addressee of the act, but granted standing based on their being individually and directly concerned by it:

as the applicant is the person who made the request under Article 13 of Regulation No 33/2008 aimed at the inclusion of flurprimidol in Annex I to Directive 91/414, the contested decision, in which the Commission refused to include that substance in the annex in question, is of direct and individual concern to the applicant.¹⁵⁹

When moving on to consider the applicants’ right to be heard, the Court found that the applicant (although not being the formal addressee of the act) was entitled to the rights of the defense in so far as the contested measure ‘refuses [their] request that flurprimidol should be included in Annex I to Directive 91/414 and, consequently, prohibits the applicant from marketing plant protection products containing that active substance’.¹⁶⁰

Therefore, because of their special procedural position as applicants, economic actors have both legal standing and the right to be heard in procedures for acts of general application. The same does not apply to other parties, even where they participated in the procedure via a public consultation on an act of general application. In PAN Europe, one of the environmental non-governmental organisations (NGOs) acting as applicants, had participated in the expert consultation envisaged by Article 12 of the Pesticides regulation. It had submitted written observations on the draft assessment report on the substance undergoing authorisation. According to the Court, however, ‘such participation does not support the conclusion that the act in question directly affects an applicant,’¹⁶¹ leading to the denial of legal standing. Likewise, participation in public consultations ‘does not grant the interested parties specific procedural rights,’ but only provides for the right to submit comments.¹⁶² Since the latter alone cannot ground legal standing nor the right to be heard, third parties other than the industry applicants will rarely be able to have the right to submit comments reviewed by the Court.¹⁶³

The effect of this case law is that economic actors are structurally privileged even in procedures of general standard-setting, in which the right to be heard, in principle, does not apply. More generally, the legal approach to standing and right to be heard is in contrast with the legislative acknowledgement of the value of public participation in risk regulatory procedures.¹⁶⁴ REACH is the flagship framework for inclusive governance in risk regulation,¹⁶⁵ explicitly envisaging consultations with stakeholders;¹⁶⁶ the Pesticides regulation provides for the already mentioned possibility of public consultations, in which environmental NGOs regularly participate.¹⁶⁷ The 2019 reform of the General Food Law has mainstreamed public consultations into EFSA’s scientific assessments.¹⁶⁸ The pharmaceuticals regulation does not expressly provide for participatory avenues, but the EMA has a well-developed system of public consultations and stakeholder involvement.¹⁶⁹ Yet while all stakeholders can participate in public consultations, only few, mostly the economic actors, can challenge the outcomes and invoke their procedural rights before the Court.¹⁷⁰

Concerning the standard of review, of the three procedural principles considered here, the right to be heard results in least intrusive scrutiny. The first scenario in which the right to be heard is invoked, is when applicants lament their inadequate involvement in the procedure. The Court generally considers how and in which context they have been invited to formulate comments.¹⁷¹ In Bayer, for example, it reviewed the decision-making process and found that

‘the applicants were invited to make comments and that they did make them, both in writing and, through the organisations representing them, at a meeting with the Commission’s services. In those circumstances, the Commission was fully entitled to regard itself as being sufficiently apprised of the applicants’ point of view and, in particular, it was not obliged to act on Bayer’s requests to meet Commission staff responsible for reviewing the substances covered’.¹⁷²

While the Court sometimes engages in a more detailed review,¹⁷³ it does not generally question the quality of the applicants’ involvement, being satisfied with the fact that such involvement took place.

The second scenario concerns the applicants’ submission of additional data to address data gaps emerged during the risk assessment. The Court consistently acknowledges the possibility for the administration to deny the admissibility of such additional data, based on the argument that the public interest in the immediate implementation of the measure prevails over the ‘applicants’ interest in having the time necessary to generate the missing data’.¹⁷⁴ The opposite solution ‘would amount to granting the applicants the right to unduly delay the adoption of a decision to withdraw or amend the approval’.¹⁷⁵ In this context, legislative objectives operate as a shield for administrative discretion. The third and final scenario is that in which, when reviewing whether the evidence submitted has been accurately considered, the Court refers cursorily to its reasoning under the duty of care and does not further engage with allegations of inaccuracy under the right to be heard.¹⁷⁶

E. CJEU review – substantively intrusive and procedurally exclusive

Our case law analysis leads to two important conclusions regarding intensity of review and interpretation of principles of administrative procedure respectively. Firstly, the three procedural principles considered above trigger different degrees of review intensity concerning technical discretion. The duty of care triggers high intensity, whereby the Court closely reviews the quality of the scientific information and reasoning following the plausibility test. Here the Court comes close to the substitution of the assessment of the agency with its own assessment. The duty to state reasons triggers a medium-level intensity, whereby the Court reviews the clarity and consistency of the reasons adduced by the Commission or the agency, not expecting them to be exhaustive but rather to focus on the decisive elements of the decision. Review under the right to be heard results in the least engagement with scientific information and reasoning.

Secondly, despite the different nature of these principles in administrative procedure, the Court interprets their function in a similar narrow way even in procedures of general rulemaking. All three play out within the bilateral relationship between the applicant as the (quasi-)addressee and the EU administration. All three are seen as compensating for the limits of substantive review and are used to protect the individual applicant from arbitrary decisions. All three principles, therefore, serve a narrow circle of (quasi-)addressees of the regulatory act, and hence in most cases the industry applicant. If the judicial protection of these principles does enhance the deliberative quality of regulatory decision-making, the result is an impoverished version of deliberation confined to the EU administration and the regulated industry. Given the epistemic imbalances characterising EU risk regulation, this amplifies the latter’s epistemic power over the regulatory process.

Our analysis sheds new light on the Court’s modern approach to reviewing risk regulation. Intensified scrutiny and the use of procedural principles to review administrative discretion remain two consistent features of the case law. The third feature we identified is the exclusivity of the Court’s approach. Instead of catalysing inclusive procedures that open regulatory science to public scrutiny, as some scholars had hoped, the Court’s approach fosters an exclusive bilateral exchange between the administration and the industry applicant. Our findings call into doubt the widely shared idea that, overall, the Court manages to strike a reasonable balance between effective judicial protection and the need to respect administrative discretion; and that it performs a ‘substantially deferential, but procedurally activist’¹⁷⁷ review. Instead, we observe a certain blurring between procedure and substance.¹⁷⁸ While the Court formally states its intention to keep procedural and substantive considerations separate,¹⁷⁹ the two often collapse into a more searching review.¹⁸⁰ As some authors have noted, this leads to the Court sometimes approaching a substantive review of the merits of scientific opinions, risking to overstep its institutional competence by behaving as a super expert, with Bilbaìna v Commission being one of the most problematic examples.¹⁸¹

This begs the question, whether the Court really respects administrative discretion, as a ‘scope for assessment and judgement left open to the decision-maker by terms of his authority’.¹⁸² There seems to be little clarity in the case law about the conceptual meaning of administrative discretion.¹⁸³ While the Court seems to distinguish different types of situations, in which discretion exists (technical assessments, factual determinations¹⁸⁴ or political assessments),¹⁸⁵ this is without consequence for the applicable standard of review, as all types of discretion are reviewed according to the manifest error of assessment standard.¹⁸⁶ Moreover, post-Pfizer the expected level of scrutiny the Court will apply in any given case is difficult to predict.¹⁸⁷ The manifest error review has changed meaning over time and is applied inconsistently, sometimes with more and sometimes with less intensity,¹⁸⁸ without the Court explaining what the actual standard is and how it arrives at the conclusion that it is (not) met. Hence the manifest error formula is rightly said to conceal more than it reveals.¹⁸⁹

The fact that the Court so far has failed to define a principled approach to the standard of review¹⁹⁰ raises doubts as to whether the Court recognises technical discretion at all. Some authors have indeed argued that there is no constitutional basis for judicial self-restraint regarding factual determinations or scientific assessments as opposed to political assessments.¹⁹¹ The Court therefore would have the last say on technical discretion and reserve the right to intervene defining the intensity of review case-by-case.¹⁹² Although we do not share this view, and although it seems to contradict the Court’s statements, it would explain the inconsistent levels of intensity in the case law.

The lack of clarity and consistency in the case law makes it difficult to formulate a convincing theory of judicial review of administrative rulemaking in the field of EU risk regulation as well as to normatively evaluate the Court’s approach. The Court rarely explicates the higher normative principles driving its review. The repeated emphasis on the need to respect administrative discretion, and to resist substitution of judgement¹⁹³ does not per se explain the normative reasons that require deference to the administrative decision-maker or that would explain differing levels of intensity in judicial review. A mere reference to the nature of complex assessments fails to justify, in normative-constitutional terms, when and why review should be limited. It has rightly been pointed out that the ‘complexity formula has become a rhetorical façade substituting a real discourse about the courts’ powers under the Treaty.’¹⁹⁴ A proper justification of judicial review of complex assessments would require a reflection on broader questions, such as on the Court’s institutional role vis-à-vis the EU administration¹⁹⁵ as well as, more generally, on the legitimacy of public administration in the EU polity.¹⁹⁶

5. The role of judicial review in expert accountability

How could the Court play a more constructive role in enhancing the accountability of EU regulatory science? In the following we argue that EU judicial review must be able to adequately respond to the concerns around epistemic capture which we identified in this paper. At present, there seems to be little awareness of this problem in the case law. It is true that the Court threads carefully and cannot be accused of siding with industry applicants. In fact, in most cases, the outcome of litigation favours the EU administration.¹⁹⁷ Yet, as we have shown, the Court’s reasoning entrenches certain ideas about legal standing and the administrative process that benefit the regulated industry, while reducing the discretion of the EU administration. Moreover, the current approach seems to ignore the reality of modern risk regulation. The latter is dominated by administrative rulemaking of high salience for health and the environment affecting a broad range of actors, which arguably should make wide access to the Court and the enforcement of epistemically inclusive procedures a priority.¹⁹⁸ The Court also ignores that the industry applicants are not merely the initiators and addressees of regulation but also key information providers.¹⁹⁹

In this section, we discuss the role of the Court from the perspective of political epistemology,²⁰⁰ which studies the role of experts in modern democracies.²⁰¹ We are particularly interested in the institutional turn in these studies, in which the question of ‘truth-sensitive’ institutional design is central. While recognising the role of politics, power, and social structure in the production of policy-relevant knowledge, this research also acknowledges the crucial role of experts in democratic problem-solving. Therefore, the focus is on the institutional mechanisms that enable expert decision-making to reconcile the ‘independence requirement of reliable expertise and the responsiveness requirement of democratic governance’.²⁰² While this literature studies the accountability of expert bodies from a broader institutional perspective, law and judicial review are increasingly seen as playing an important part in the mix of institutional mechanisms holding expert bodies to account. Courts can contribute to addressing the epistemic concerns around the use of specialised expertise in regulation.²⁰³ In other words, they can hold experts to account for the epistemic quality of their decision-making, which in turn is crucial for both upholding the rule of law and democratic problem-solving.

A particular concern in this regard is the risk of cognitive error and bias in the production of policy-relevant science, which is closely interlinked with the risk of a minoritarian bias, discussed at the beginning of this paper. Studies show that experts can make cognitive errors, are prone to over-confidence, and can suffer from cognitive or ideological bias.²⁰⁴ In fact, scientific experts have been shown to suffer from the same biases as lay people and moreover show ‘a strong affiliation bias, with those experts working in industry having a more benign view of the relevant risks that those experts working outside it’.²⁰⁵

Bias can also result from special interests distorting the production of high-quality expert advice. This does not have to be through direct manipulation of scientific data or findings, but in a more subtle way. As Christiano puts it

It seems to me that one principal source of domination of and parochialism in science is domination by class or ethnic interests. The reason for this is based in the simple facts of cognitive bias. Those facts suggest that beliefs and science can easily come to reflect the narrow backgrounds and interests of those who produce them. So, if science and expertise are funded by one particular group in the society there is a significant danger that that expertise will reflect the interests and backgrounds of those persons.²⁰⁶

As we explained earlier, EU risk regulation is not immune to such concerns, as the industry applicants fund and provide most of the evidence based on which market authorisations are granted. EU agencies review such evidence and are supposed to do so in an independent way and considering other scientific sources. However, there are ongoing concerns about both the independence and quality of agency assessments,²⁰⁷ which moreover take place within often underfunded institutions.²⁰⁸

There are several institutional mechanisms to tackle parochial and biased production of knowledge.²⁰⁹ Some scholars differentiate between mechanisms targeting expert behaviour, expert judgement, and the conditions for expert inquiry. ²¹⁰ Others distinguish between ex-ante and ex-post mechanisms of accountability.²¹¹ In one way or another, all these mechanisms aim at extending expert accountability obligations from peer-accountability (peer-review among equally qualified experts in epistemic communities) to public accountability, ie to an obligation to explain and justify expert judgements to public fora.²¹²

In a process of democratic decision-making, the testing of judgments and arguments must also be extended for epistemic reasons to other relevant disciplines and other relevant expert fora, for example, to fora comprising bureaucrats and competent stakeholders, to the legislature and even to the public sphere at large. In all these fora, experts can be asked to account for critical assumptions, explain models used, specify their limits and present alternative models.²¹³

Such an obligation is justified with the need for cognitive or epistemic diversity, the latter being an important pre-condition of successful deliberation, and thus of high epistemic quality of expert decision-making. Deliberating groups are less prone to cognitive bias, group think or distortions based on conflicts of interests, enlarging the pool of ideas and information on which scientific advice is based.²¹⁴ This emphasises the importance of epistemic pluralism within the EU administrative process. The quality of EU risk assessments crucially depends on epistemic diversity and the input of all relevant expertise. In fact, these ideas are already reflected in most of the EU risk regulatory frameworks²¹⁵ and in many EU soft law documents.²¹⁶ If deliberation and pluralism are recognised values of the EU regulatory process, this should, in our view, also be reflected in the procedural review of that process by the CJEU.

This brings us back to the Court’s role in ensuring expert accountability. We propose that judicial review can be an important mechanism of expert accountability, in two ways. Firstly, courts directly provide a public forum, in which expert judgements are reviewed. This is what the CJEU review of risk regulation already does, when reviewing the reasonableness of science-based decisions against the background of the information submitted by the litigant challenging those decisions. However, while a useful starting point, this pathway is not sufficient for meaningful expert accountability. As we have seen, by narrowing down the circle of persons who, as litigants, can bring counter-expertise to challenge EU regulation, the Court’s approach fails to strengthen epistemic diversity and deliberation. Instead, it entrenches the position of the regulated industry as a dominant information provider throughout the regulatory and litigation cycle. Moreover, for reasons of institutional balance and required technical expertise, we believe that judicial review of regulatory science should in principle be limited and becomes problematic where the Court oversteps the boundary of procedural reasonableness review.

Therefore, secondly, a more promising avenue is for judicial review to indirectly improve expert accountability within the regulatory process. This is where our work connects to the ‘courts as catalysts’ literature, which suggests that beyond legality review, courts should also fulfil a quasi-regulatory function by promoting the adherence to certain procedural standards which enhance the legitimacy of the regulatory process.²¹⁷ The idea of a catalyst function of judicial review chimes in with the insights on expert accountability presented above. It also places the issue of expert accountability within the broader discourse on the role of deliberation in democratic decision-making and corresponding conceptions of the role of judicial review.²¹⁸ According to such conceptions, the CJEU can help ensure deliberation as an important legitimising principle of EU governance, deliberation being understood as a process in which ‘a wide variety of actors are encouraged to come forward to present reasons and interrogate those of others, with a view to solving problems and providing norm guidance after an exchange of opinions and views.’²¹⁹ The Court’s role should be to uphold standards that ensure that decisions are reasoned (ie are justified by reference to reasons acceptable to those affected by those decisions); oriented towards problem-solving by allowing for contestation and divergent views in the decision-making process; and preceded by participation of all affected actors.²²⁰

In our view, embracing such normative ideas in CJEU review of risk regulation would help address the risk of epistemic capture. However, in contrast to some of the previous literature on this topic, we do not find evidence of the Court embracing these ideas in the risk regulation case law over the last decade.²²¹ Instead, we show that as long as the procedural values of duty of care, reason-giving and participation (ie the right to be heard), are not interpreted with the value of epistemic diversity and inclusive deliberation in mind, they fail to catalyse meaningful expert accountability mechanisms.

6. What way forward for CJEU review of regulatory science?

The above formulated normative expectations towards judicial review raise the question whether we are possibly expecting too much from the Court. On the one hand, the mandate of the Court is stated to be one of ‘observing the law,’²²² which could be seen as supporting the focus on legality in the case at hand rather than broader institutional questions. On the other hand, even if one agrees that judicial review can and should have broader catalytic effects on governance processes,²²³ one should perhaps be careful to place responsibility for institutional change on the Court’s shoulders alone. After all, other governance actors, above all the EU legislature and the EU administration, are responsible for ensuring a legitimate administrative process,²²⁴ which is conducive to producing excellent, independent, and transparent advice. It could be argued that it is up to those other institutional actors to provide for stronger participation rights and to ensure epistemic diversity is safeguarded. The Court acts within the parameters set by those institutions.

While these arguments are valid, we see room for improvement. First, legal norms in risk regulation are often broadly formulated with their legal determination depending on scientific arguments. Legal and scientific rationality are enmeshed. Therefore, ensuring that the law is observed in areas of high techno-scientific complexity requires mechanisms of expert accountability, to which the Court should contribute. Second, as we discussed above, judicial review is an important mechanism for ensuring the high epistemic quality of regulatory science, next to other institutional mechanisms. Third, we acknowledge that the Court should contribute to EU expert accountability not in isolation, but together with other actors. This means that legislative and administrative action will at times be required before the Court can fully unfold its catalyst function.

A comparison to the CJEU case law on transparency and access to documents provides a useful example.²²⁵ Here the Court, relying on a clear legislative mandate, has been able to strengthen access rights vis-à-vis regulatory science, at least around access to environmental information.²²⁶ On the one hand, applicants who are denied access to information are expressly granted standing,²²⁷ without having to fulfil the Court’s test under Article 263(4) TFEU for regulatory acts. As a result, public interest groups, as well as individuals, can challenge the EU administration’s decisions granting confidentiality to industry studies. This leads to a significantly more balanced litigation pattern, whereby 10 out of 19 cases over the period of 10 years (2010–2020) have been initiated by public interest applicants, the majority of which (7) successfully. On the other hand, following the wording of the Transparency regulation’s recitals, the Court explicitly interprets transparency in light of epistemic diversity, public participation, and the democratic legitimacy of EU institutions.²²⁸ This has led to new judicial approaches to transparency, in particular in the field of environmental protection.²²⁹ The contrast between access to documents litigation and risk regulation litigation as discussed in this paper creates a paradoxical result: public interest groups and citizens can go to court to enforce their access to the scientific information held by EU agencies, but they do not have standing to challenge decisions based on that information.

This example shows that the Court can catalyse better procedures in interaction with the legislature and the administration.²³⁰ Choices made by other EU actors therefore influence how far the Court can go in strengthening certain procedural values.²³¹ Participation rights in the administrative process are especially sensitive. It could be argued that the Court is mindful not to encroach upon the powers of the legislature, and therefore refuses to provide participation rights beyond the rights of defense in individual procedures.²³² Participation rights for acts of general application, as those often found in risk regulation, should therefore be grounded in legislation. EU legislation, however, currently does not provide robust and enforceable participation rights beyond the weaker right to be consulted in public consultations. While regulatory frameworks provide for consultation of all stakeholders, enforcement of consultation rights is reserved for individuals who have standing, because they are directly and individually concerned by the contested measure, namely the industry applicants.²³³

Absent legislative intervention, what could the Court do to enable a more meaningful expert accountability and to address the epistemic power imbalances in the production of EU regulatory science? The gist of our proposal is twofold: On the one hand, we believe that there is room for a different, more generous judicial interpretation of legal standing as well as of the administrative principles discussed in this paper. Such interpretation should embrace the values of inclusive deliberation and epistemic diversity thereby overcoming the current exclusive focus on the bilateral relationship between the applicant and the administration. On the other hand, if the Court maintains its current narrow interpretation, more deference and a less intrusive review of technical discretion are warranted.

Within the scope of this paper, we are not able to comprehensively discuss how exactly the Court’s new interpretation of legal standing and administrative procedure should look like. The aim of our analysis has been to reveal the ways in which the CJEU doctrine interacts with the epistemic power of economic actors with potentially problematic consequences for the epistemic quality and public interest orientation of EU regulation. However, beyond critique we would also like to point to normative approaches that might counteract the problematic dynamics, which we identified in this paper. We are moreover aware of the doctrinal complexity of administrative legal concepts as developed in EU case law, which demands a more thorough analysis of our proposals than we can offer in this final outlook. Such analysis will have to be reserved to future research building on existing EU administrative law scholarship.

We have shown that to counter the epistemic power of the regulated industry, the Court’s approach to both legal standing and procedural review should change. While the legal standing test privileges the regulated industry as a challenger of regulation, the current interpretation of procedural principles enables, to varying degrees, an intensified scrutiny of the scientific basis of regulation based only on claims brought by the industry.

Concerning legal standing, the Court’s approach to direct and individual concern under Article 263 (4) TFEU, justified by the famous, yet often criticised, reasoning that the EU’s system of judicial protection is based on a complete system of remedies,²³⁴ seems to be too entrenched to expect the Court to significantly change course. As discussed in Section 3, the Court has maintained its approach despite widespread and persistent criticism. More recently, in the light of the climate emergency, such criticism has gained even more weight in the field of environmental protection.²³⁵ The Court’s narrow and individualised understanding of the circle of affected interests seems to significantly clash with the collective nature of environmental harm. It also falls behind the ambition of the EU’s Green Deal of a society-wide green transition. At the international level, the Aarhus Convention Compliance Committee has recently declared the EU to be in violation of its international commitments concerning access to justice in environmental matters. None of these arguments, we believe, are likely to change the Court’s position, which is based not only on a formalistic understanding of the Treaty text, but also on a deep-seated systemic understanding of the interplay between EU and national courts as well as on practical considerations – never explicitly stated – pertaining to issue of case load and the functioning of the Court.²³⁶ While we support a broader interpretation of legal standing at least in the field of environmental and health protection, we acknowledge that change is more likely to come from EU legislature.²³⁷

Assuming access to litigation remains restricted, how much would be gained from a re-interpretation of the rights of administrative procedure? And would it be possible given that the Court’s understanding of who may challenge administrative acts is directly linked to the question of who may derive such procedural rights? In the following, we differentiate between the right to be heard on the one hand, and the duty of care and reason-giving on the other hand, for two reasons. Firstly, in the absence of change on legal standing, a broader interpretation of the right to be heard is not only less likely,²³⁸ but also less impactful, because it will continue to benefit a small circle of industry applicants. Secondly, in the case law the invocation of the right to be heard does not lead to intensified review in contrast to the invocation of the other two principles. The Court’s interpretation of the duty of care enables the most intensive review of technical discretion. Therefore, under the current institutional framework, the most impactful change would be a re-interpretation of the principles of the duty of care and of reason-giving. At the same time, from the Court’s perspective, a broader interpretation of both principles might be easier to embrace as compared to a change of course on the right to be heard.

To begin with the duty of care, this principle has not only been key in enabling more judicial engagement with questions of epistemic quality but is also a procedural guarantee with a strong public interest dimension,²³⁹ which resonates with ideas of inclusive deliberation and epistemic diversity. While the principle of duty of care can sometimes be seen as a rule protecting individual interests,²⁴⁰ its primary function is that of an objective procedural guarantee arising from an absolute obligation on the part of the EU administration to examine carefully and impartially all the relevant aspects of the individual case. This is particularly the case for general rulemaking,²⁴¹ where its protective function for individuals diminishes as compared to individual decisions. Scholars have emphasised the potential of this principle to induce more deliberation in administrative procedures concerned with the gathering of highly technical yet publicly salient information.²⁴² Nothing in the Court’s reasoning indicates that there are principled reasons to reduce the scope of relevant information to be duly considered by the administration in the process of scientific assessment to only the information submitted by the applicant. Rather, such reduction currently happens in the case law because the Court responds to the arguments brought by the litigant. It is thus an indirect result of the legal standing requirement. It is also a result of the Court’s emphasis that the duty of care compensates for limited substantive review and thereby protects the interests of the applicant. Yet in contrast to the right to be heard, in the Court’s reasoning, the duty of care does not depend on the test of individual and direct concern. Indeed, it is understood that other actors participating in the regulatory process, eg third parties via the public consultation procedures, are also providing relevant information, which, if prompted, the Court would consider.²⁴³

An example from the Court’s case law on tobacco control may help clarify our argument. In Pillbox the Court decided a request for a preliminary ruling challenging the validity of EU legislation on electronic cigarettes.²⁴⁴ The main disputed legal issue pertained to the quality of the scientific basis of the legislation as part of the proportionality test. More specifically, it was contested whether there was sufficient evidence of health risks arising from e-cigarettes that would justify the introduction of restrictions. The way in which the Court responded to the challenge of the scientific basis of the Directive mounted by the industry before the national court was by examining the quality not so much of the evidence as such, but of the process whereby the EU legislature collected such evidence. Before adoption, EU legislature undertook extensive consultations with experts and concerned parties to consider all information. The Court recognised that this did not remove scientific uncertainty. Nonetheless, the fact that the European Parliament, the Council and the Commission ‘took account of the available scientific evidence and the opinions of the interested parties’ and that ‘a number of consultations and meetings was organised precisely in order to collect the necessary information on the options available to the EU legislature’ including industry and consumers representatives, scientists, national and international authorities,²⁴⁵ allowed the Court to conclude that the measure rested on a sound epistemic basis. By emphasising the inclusiveness of the consultation process the Court implicitly emphasised the value of epistemic diversity in the legislative process, which also served as a defense against the applicant’s challenge amidst scientific complexity.

The Pillbox approach to procedural review²⁴⁶ has no equivalent in relation to administrative acts, albeit of general application, such as those considered in this paper. Despite the different legal context, we believe that an analogous approach should apply when the Court applies the principle of the duty of care in administrative rulemaking. At present, in industry-led litigation, the Court confines its review to inquiring whether the information invoked by the applicant was duly considered in the administrative process. We suggest that, in the light of the values of epistemic diversity and the objective procedural nature of the principle of duty of care, the Court should instead look at how inclusive the process of information gathering was, and whether the administration considered all relevant information, including those submitted by other stakeholders.²⁴⁷ Of course, considering the current litigation patterns outlined in Section 3, it would mostly not be prompted to do so by the litigants but would have to do so ex officio. The application of the duty of care in state aid cases indicates the existence of an obligation to look beyond the issues raised by the applicant.²⁴⁸ From an external perspective, this would allow the Court to alleviate the epistemic asymmetries in EU risk regulation. Judicial review could thus catalyse procedures in which agency experts explain and justify their judgements not only vis-à-vis the regulated industry, but also vis-à-vis other stakeholders and affected actors. From an internal perspective, this could serve as a normative benchmark against which to measure the intensity of judicial review. The finding that the administration gathered and duly considered information from different sources and stakeholders should justify less intensive review and more deference – the lack thereof a more intensive engagement with the scientific information.²⁴⁹

A similar understanding should inform the reason-giving requirement, which, as we discussed, also has a strong objective procedural dimension.²⁵⁰ It serves as ‘an indicator of substantive defects in the decision to be examined’.²⁵¹ The requirement of reason-giving flanks the application of the principle of duty of care,²⁵² both reinforcing the reasoned nature of the decision-making and ensuring that all the relevant circumstances were duly considered in the decision-making process. Reason-giving, therefore, should be directed at a broader range of stakeholders, refer to all relevant information submitted by them, display all relevant arguments and how they have been weighted against each other, and how disagreement has been handled.²⁵³ At present, however, the Court understands the requirements of reason-giving to be a corollary to the right to be heard, understood as a right of defense of the ‘adversely affected individual’.

In the absence of legislative or interpretative change, the normative value of the Courts’ intensified review of technical discretion, as developed post-Pfizer, should be questioned.²⁵⁴ As long as only a small group of powerful economic actors has access to litigation and benefits from process rights, such review tends to aggravate the epistemic power imbalances and related legitimacy concerns affecting EU risk regulation.

7. Conclusion

Post-Pfizer, the CJEU approach to reviewing both administrative discretion and administrative process has often been understood as a welcome response to a maturing and expanding EU administrative state. The two interlinked features of the modern approach – ie intensified scrutiny of the scientific basis of administrative decisions and procedural review based on the principles of duty of care, reason-giving and participation – promised to achieve a Goldilocks balance between the Court’s task of ensuring judicial protection amidst scientific complexity, on the one hand, and its duty to respect the administration’s prerogatives as decision-maker, on the other hand. Moreover, the Court has been described as an ‘information catalyst’ promoting inclusive procedures which generate deliberation among all relevant experts, thereby enhancing the informational basis behind EU regulatory measures.

In this paper, we have questioned these established accounts offering a critique of the case law since Pfizer on both doctrinal and socio-legal grounds. From a doctrinal perspective, effective judicial protection is seriously undermined by the lack of consistency and conceptual clarity in the case law. Moreover, the blurring of substantive and procedural review and the resulting intensification of scrutiny of scientific information calls into question the principle that in situations of high techno-scientific complexity the administration enjoys a wide discretion.

From a socio-legal perspective, we drew attention to the often-overlooked role of economic actors as information providers in both the EU regulatory process and litigation; as well as to the ways in which CJEU doctrinal approaches interact with the epistemic power of economic actors. We have shown that economic actors exert epistemic power over EU regulation, both through lobbying and as key information providers via legislative design. By feeding information to regulators, economic actors shape the epistemic basis of EU risk regulation, giving rise to the risk of a minoritarian bias, ie the industry’s over-representation in the assemblage of scientific knowledge underpinning regulation, and particularly administrative standard-setting.

Such risk of overrepresentation exists also at the stage of litigation due to the CJEU judicial interpretations of legal standing and administrative procedural guarantees. For both access to the Court and procedural guarantees, the Court significantly reduces the circle of beneficiaries to actors who as (quasi-)addressees and procedural applicants can claim a special bilateral relationship with the EU administration. The Court also reduces the function of principles of administrative procedure to a protective function whereby they serve to compensate for the limits of substantive review and to protect the individual applicant from arbitrary decisions. All three principles, therefore, serve a narrow circle of (quasi-)addressees of the regulatory act, and hence in most cases the industry applicant. This narrow compensatory understanding applies also to procedures of general rulemaking, thereby undermining the objective public interest nature of the principles of duty of care and reason-giving.

Therefore, instead of catalysing inclusive procedures that would open regulatory science to public scrutiny by a wide range of actors, as some scholars had hoped, the Court’s approach fosters an exclusive bilateral exchange between the administration and the industry applicant. Care, reasoned justification and right to be heard are owed primarily to the industry applicants. Thus, the Court scrutinises the scientific basis of administrative rulemaking from the vantage point of resourceful economic actors. Seen together with their epistemic influence in EU regulation and a legal doctrine on standing that grants economic actors almost exclusive access to litigation, the Court’s modern approach fails to address, and arguably further entrenches, the epistemic power imbalances inherent in EU risk regulation.

What is the way forward for CJEU review of EU regulatory science? We have argued that the Court should play a more constructive role in enhancing the accountability of EU regulatory science by paying more attention to the risk of epistemic capture. There is, we believe, room for a different, more generous judicial interpretation of legal standing as well as of the administrative principles discussed in this paper. Given that the Court’s approach to legal standing and to participation rights (both being closely intertwined) is unlikely to change soon, we argue that under the current institutional framework, the most impactful change would be a re-interpretation of the principles of the duty of care and of reason-giving, which currently lead to the most intrusive level of review. In our view, the Court should stronger emphasise the objective procedural nature of these principles when reviewing the scientific basis of EU general rulemaking. In so doing, it should embrace the values of inclusive deliberation and epistemic diversity thereby overcoming the current narrow and exclusive focus on (quasi-)addressees. This would allow the Court to look beyond the claims of the applicant and to consider the overall inclusiveness of the process of information gathering. Instead of only scrutinising whether the EU administration has duly considered the information submitted by the industry applicant, the Court should also scrutinise, ex officio, whether the administration has considered all relevant information, including those submitted by other stakeholders. Judicial review would thus catalyse procedures in which agency experts explain and justify their judgements not only vis-à-vis the regulated industry, but also vis-à-vis other stakeholders and affected actors.

In the absence of either legislative or interpretative change regarding access to the Court and rights of administrative procedure, more deference and a less intrusive review of technical discretion are warranted. The intensified standard of review as developed post-Pfizer is open to critique both because it lacks proper doctrinal justification and because of its problematic socio-legal effects. As long as only a small group of powerful economic actors has access to litigation, and benefits from procedural guarantees, such review risks aggravating the epistemic power imbalances and related legitimacy concerns affecting EU risk regulation at present.