14 - Soft Law and Challenges to Access to Justice

Summary

This chapter illustrates how soft law can have very real implications for the fundamental rights position of individuals. While the existence of an interference is often treated implicitly in the case law of the ECtHR and the CJEU, the chapter argues that interferences can result from soft law acts. For the EU this means that recourse to soft law does not in itself preclude the application of the Charter. How soft law may concretely interfere with fundamental rights is illustrated with a couple of examples from different policy areas. Having shown the possible fundamental rights implications of soft law, the chapter turns to remedies. Here the limitations of the EU’s system of judicial remedies is highlighted and the potential of non-judicial remedies is explored.

14.1 Introduction

This chapter explores the challenges that the informalisation of EU governance, through the reliance on soft law, poses for the possibilities for private parties to challenge EU conduct that may be in violation of their fundamental rights. First, this main research question will be situated in the broader debate on reliance on soft law in EU law. Subsequently, the ontological question of how soft law, given its non-binding nature, may constitute an interference with a fundamental right will be addressed and different examples of the increased reliance on soft law will be used to illustrate recent developments. Having sketched the state of play and the challenges that soft law may pose for fundamental rights, the chapter will then turn to the EU’s system of remedies. The latter has been crafted with a view to providing legal protection against binding governmental action. Soft law undermines that assumption and the difficulty in challenging it through conventional mechanisms of legal protection may itself become a possible interference, this time with the right to an effective legal protection as enshrined in Article 47 of the Charter of Fundamental Rights of the EU (CFR, ‘the Charter’).¹ After having identified the (conventional) remedies available, the chapter will make some suggestions to adapt the system of remedies addressing the specific challenge posed by soft law.

14.2 Soft Law: What’s in a Name

The final part of this second section of the chapter is devoted to illustrating, through a series of examples, how, in practice, soft law could interfere with fundamental rights. Before presenting these examples, however, some preliminary clarifications need to be made. First, Senden’s definition of soft law will be presented together with a discussion of the different functions that soft law may fulfil, before dealing with the question of how, in theory, soft law may interfere with fundamental rights.

14.2.1 What Is Soft Law and What Makes It Soft?

The study of soft law in the EU legal order is already well established. A seminal work, and one that has advanced an often relied on definition of the notion, was that of Senden. She qualifies soft law as ‘rules of conduct which are laid down in instruments which have not been attributed legally binding force as such, but nevertheless may have certain (indirect) legal effects, and that are aimed at and may produce practical effects’.² This definition remains rather vague, however, and a lot of intellectual effort has since gone into attempts to better understand what kind of practical and/or indirect legal effects are sufficiently relevant in order for non-binding acts to be considered legally significant enough to constitute soft law.³ Further research has therefore looked into the question of which precise features makes soft law soft and the types of soft law that exist.

On what makes soft law soft and trying to better capture the notion of soft law, Terpan proposed to understand the phenomenon as a function of two criteria, those being obligation and enforcement. A norm can then be qualified as hard law when it cumulatively prescribes a sufficiently clear obligation that can also be enforced. Conversely, norms that do not prescribe sufficiently clear obligations would constitute soft law as would norms that do prescribe obligations but that are not enforceable. To determine whether a sufficiently clear obligation is being prescribed, Terpan considers the source and the content of the norm in question and, in terms of enforcement, he distinguishes between norms that are subject to judicial (or at least very constraining) control, those subject to non-coercive control (such as mere monitoring), and those subject to no control.⁴

14.2.2 Soft Law’s Different Functions

To better appreciate the dynamics (and interference with fundamental rights) that may flow from soft law, it is useful to distinguish the different functions that soft law acts may fulfil. Here, different authors have elaborated different typologies. Senden and Van den Brink, for instance, distinguish between soft regulatory rule-making and soft administrative rule making. The first has a para-law policy-steering function while the purpose of the other is to give post-legislative guidance.⁵ The classification of soft law adopted specifically for EU agencies as proposed by Rocca and Eliantonio also categorises soft law in light of its position in the policy cycle and can also be integrated in that of Senden and Van den Brink. Rocca and Eliantonio distinguish four types: where soft law is adopted in the form of acceptable means of compliance or technical guidance (categories one and two of Rocca and Eliantonio), it will typically fulfil the function of post-legislative guidance; conversely, technical documents and high quality information (categories four and three of Rocca and Eliantonio) typically fulfil a para-law function.⁶ Just like the origin of a norm’s soft nature above, the conceptual distinction between these two main categories is important because it feeds into the constitutional assessment of the informalisation of EU governance through soft law.

On the one hand, post-legislative soft law may be beneficial to clarify the meaning and scope of legislative provisions. On the other hand, there also exists a risk that in the post-legislative phase, through the soft law guidance, a binding norm is given (slightly) different meaning, thereby also possibly altering fundamental rights interferences flowing from that binding norm. In that case, however, a binding norm still exists that might meet the requirements of Article 52(1) of the Charter. In contrast, soft law fulfilling a para-legislative function does not have such a binding norm to fall back on, raising the question of whether any fundamental rights interferences by such soft law would be ‘prescribed by law’ or whether they would be ipso facto violations.

A final consideration to present at this point is the dual question of the author and the addressee of the soft law. In EU law, perhaps the archetypical form of soft law are the Commission’s guidelines in the field of competition law where the Commission set out its view on how EU competition law is to be applied. This soft law has no specific addressees but helps private parties anticipate the Commission’s practice, since the soft law has self-binding effect on the Commission as the (main) actor responsible for the enforcement of EU competition law.⁷ More recently, a different trend can be seen, as will be further illustrated below, whereby a proliferation in the authors of soft law can be witnessed. In this fairly new constellation,⁸ the rather straightforward link between the author adopting the soft law and the enforcement actor has also been severed, meaning there is no self-binding effect anymore either. Often EU agencies or bodies are called on to adopt soft law, fleshing out legislative provisions that are to be applied by national authorities. While such soft law may still fulfil the clarification function of post-legislative soft law noted above, the fact remains that the fiction of an authority limiting its own (enforcement) discretion can no longer be upheld in these cases. That would mean that the position of private parties is weakened, and the possible interferences with their fundamental rights may become more acute. This since the soft law will still carry in it an expectation of compliance but without being balanced by the legitimate expectation that the enforcement actor has limited its own discretion.

This feature is exacerbated by a further feature of the typical composition of EU agencies or bodies adopting soft law. Unlike the main EU executive (the Commission), the main decision-making entities of EU agencies and bodies are the national authorities responsible for the implementation of EU law. The further risk this poses is illustrated nicely by the recent ThyssenKrupp cases. At issue here was a composite procedure whereby the EU customs expert group (composed of national authorities) could not agree on whether certain conditions to grant an exceptional authorisation to one of ThyssenKrupp’s competitors were met. Still, a majority in the expert group was in favour of granting the authorisation and the Commission concluded that the conditions to do so were indeed met. The Dutch customs authority subsequently formally granted the authorisation and claimed that it was obliged to do so in light of the Commission’s conclusion.⁹ This because a further soft law act adopted by the expert group, ‘the new administrative practice’, provided that such Commission conclusions are binding on the national authorities.¹⁰ The General Court (GC) ruled that the Commission’s conclusion was not challengeable pursuant to Article 263 TFEU¹¹ since the relevant legislation merely prescribed that national authorities ‘shall take it’ into account.¹² The GC’s findings were upheld by the Court of Justice, which furthermore remarked that the ‘fact that … the competent customs authority took its decision in the belief that it was bound by the contested conclusion does not … have the effect of making that conclusion an act which is legally binding’.¹³ For readers versed in the admissibility requirements of Article 263 TFEU procedures, the Courts’ findings will appear logical and sound. However, a troubling dynamic at play here is that national authorities come together at EU level and de facto limit their own discretion through soft law, obfuscating the available remedies available to private parties. As Eliantonio explains in Chapter 13, where the EU adopts preliminary measures and national authorities adopt the final act, a disentangling of who does what becomes paramount.

14.2.3 Soft Law Interferences with Fundamental Rights in Theory

Having further clarified what makes soft law soft and how soft law may occupy different positions in the policy cycle, it is now possible to look into the implications of soft law from a fundamental rights perspective. Two questions are especially relevant here: First, if soft law is soft because it does not prescribe a hard obligation and/or because any possible obligation prescribed will not be enforced through a ‘constraining’ enforcement mechanism, can it become sufficiently relevant for its fundamental rights implications? Secondly, if a limitation or restriction may result from soft law, under which conditions may those limitations or restrictions be said to be ‘prescribed by law’? Under the Convention, this requirement first refers back to the domestic law of a party but that domestic law must itself also be in conformity with the Convention. While the domestic law need not formally be law, and can be enactments of lower rank than statutes and even unwritten law, it must meet qualitative requirements, including those of accessibility and foreseeability.¹⁴ While this seems to provide sufficient flexibility to encompass soft law, the flexibility is arguably aimed at accommodating the different legal systems of parties to the Convention.¹⁵ Given the risks that soft law poses for executive overreach in the EU,¹⁶ limitations flowing exclusively from soft law might have to be ruled out as being ‘prescribed by law’ regardless of whether the soft law was accessible and its application foreseeable. Most of the time, however, a limitation will not flow exclusively from soft law, the latter instead fulfilling a mediating function. Where soft law is tolerated in such scenarios it becomes crucial to ensure proper remedies, as discussed in Section 14.3.

Returning to the first question, the relevance arguably lies in the practical and indirect effects that soft law may produce. After all, from a fundamental rights perspective, the question is not so much whether the norm prescribing (in)action on the part of the government is valid and/or binding but if and how such (in)action impacts the fundamental rights position of the individual.

The constitutional relevance of soft law is still obscured, however, since the notion of a limitation or restriction (of a fundamental right) has arguably not been properly clarified by either the European Court of Human Rights (ECtHR) or the Court of Justice of the EU (CJEU). For the ECtHR, this might not come as a surprise since the cornerstone of the Convention is the right of individual petition.¹⁷ The ECtHR thus addresses specific (individual) complaints of concrete (alleged) violations of fundamental rights and not abstract questions relating to possible dubious legislative acts or government policy.¹⁸ Such a constellation provides less opportunity for theory-crafting compared to the system in which the CJEU operates of which the purpose is not to protect fundamental rights but to oversee the construction of an internal market. In its system of remedies, the cornerstone is the preliminary reference procedure¹⁹ and an independent Commission can bring infringement actions against Member States. Returning to the ECtHR, Letsas notes that its diagnostic test has five main stages, those being: (i) whether the facts come within the ECHR’s scope, (ii) whether there is an interference with a right, (iii) whether the interference is prescribed by law, (iv) whether a legitimate is aim pursued, and (v) whether the interference is necessary in a democratic society.²⁰ While the second stage is conceptually distinct from the others,²¹ in practice it is often subsumed under the fifth stage, which has also meant that the ECtHR has not developed a horizontal approach to the question of which types of governmental action or inaction may amount to an interference.²² Since such an approach is lacking, the ECtHR has not articulated a general approach to the potential fundamental rights implications of soft law either.

When the interference question is explicitly addressed by the ECtHR, it is necessarily done in a case-by-case manner. In those cases, it is not exceptional for the ECtHR to simply postulate that there has been no restriction, without elaborating any reasoning in this regard. It thus, for instance, held that ‘in themselves, the security checks to which passengers are subject in airports prior to departure do not constitute a restriction on freedom of movement’.²³ On other occasions, the finding of a lack of a restriction is more reasoned. In Cha’are Shalom Ve Tsedek v France, the ECtHR held that while ritual slaughter was protected under Article 9 of the Convention, ‘there would be interference with the freedom to manifest one’s religion only if the illegality of performing ritual slaughter made it impossible for ultra-orthodox Jews to eat meat from animals slaughtered in accordance with the religious prescriptions they considered applicable’.²⁴ Similarly, in A, B and C v Ireland, the Grand Chamber of the ECtHR ruled that there was no interference with a woman’s right to life by a party’s anti-abortion legislation when the woman in question could go abroad to receive treatment.²⁵ As Gerards notes, if the Court were to follow this approach consistently, not many interferences would be found,²⁶ since an interference could only be qualified as such when it resulted in the impossibility to enjoy the protection of the right concerned. Linking back to soft law, that threshold might never be met since its soft nature means that non-compliant behaviour remains entirely possible.

On the other hand, in a greater number of cases, the Court links the degree of interference with the intensity of the proportionality review.²⁷ As a result, ‘although in some cases the Court pays express attention to the phase of interference, in other cases it implicitly accepts or assumes the existence of an interference, or it merges the test of interference with the test of applicability or that of jurisdiction [or that of the justification of an interference]’.²⁸ Linking back again to soft law, this would mean that the existence of an interference will be accepted but that the proportionality review might be lenient. The lack of a clear position on the possible interfering effects flowing from soft law contrasts with the approach of the CJEU, at least in the area of the internal market.²⁹ As noted, this can arguably be explained by the different purposes of both Courts and the different procedural avenues to reach them.

In its internal market jurisprudence, it is firmly established case law that where national measures ‘are capable of hindering or rendering less attractive the exercise of the fundamental freedoms guaranteed by the Treaty’ they will constitute interferences.³⁰ That non-binding measures may also constitute restrictions on free movement is explicitly established in case law of the CJEU, since the restrictive effect of non-binding measures may be comparable to that of binding measures.³¹ The degree of interference will then only be relevant in the possible assessment of the proportionality of the (national) measure. Of course, this ‘internal market’ logic might not be transposable to the fundamental rights protection offered by the Charter. While the CJEU itself has created a bridge between the two, that bridge does not connect the internal market freedoms with the full scope of material rights in the Charter. Instead, the Court held that where there are interferences with the fundamental freedoms of the internal market, these automatically cover interferences with only a limited subset of Charter rights (notably those laid down in Articles 15 to 17),³² amalgamating the assessment of both. Going back to soft law’s effects and as AG Bobek explained, soft law as an imperfect norm may not be coercively enforced but still have the ‘normative ambition of inducing compliance’.³³ In the law of the internal market, such an ambition is sufficient for the measures to be caught by the free movement rules. Applying the same logic, the possibility of soft law constituting an interference with fundamental rights should be accepted.

The more concrete question of whether this means that the Charter applies to EU soft law further falls apart in two sub-questions, in light of the Charter’s field of application. As Kenner and Peake note, the first is whether EU institutions are bound by the Charter when adopting soft law. For those authors, ‘ideally, when it adopts non-binding policies, it should also be regarded as constrained and empowered by fundamental rights. However, in those contexts, as its output is not legally binding, it would be difficult to enforce this commitment in the courts’.³⁴ The second question is whether EU Member States are to be considered as implementing EU law when they adopt binding measures in the wake of EU soft law. As the law currently stands, Kenner and Peake answer this question in the negative but again call for a de facto respect for the Charter in order not to erode the protection offered by the latter.³⁵ At this point, it is useful to make a distinction between the different ways in which EU soft law may be followed upon or implemented. Thus, once an EU institution or body adopts soft law, generally four scenarios are possible: (i) the soft law is addressed to private parties (or has no addressee) and is acted upon by them; (ii) a governmental body (EU or national) adopts a binding act further to the EU soft law; (iii) a national body adopts soft law further to the EU soft law; (iv) the EU soft law is acted upon through factual conduct of a governmental body (EU or national). Evidently, the possibility for parties to bring a potential fundamental rights interference before a court will be influenced by which of the four scenarios is at issue.

While the Court has not addressed the first scenario explicitly yet, the dynamics that would be in play have popped up in cases before it. In Bevándorlási és Állgmpolgársági Hivatal, the Court noted that asylum seekers cannot be required to undergo a personality test, since their consent to do so would be required. However, it explicitly accepted that ‘that consent is not necessarily given freely, being de facto imposed under the pressure of the circumstances in which applicants for international protection find themselves’.³⁶ The Court thus recognised that an applicant might be indirectly or practically forced, in which case there is an interference with the right to respect for a private life that needs to be justified. In the separate context of EU-induced national austerity programmes, which come under the second of the four scenarios noted above, AG Saugmandsgaard Øe in Associação Sindical dos Juízes Portugueses held that the Council recommendation based on Article 126(7) TFEU at issue ‘did not fix sufficiently specific and precise objectives to support the view that the Portuguese State implemented on the basis of that recommendation requirements of EU law within the meaning of Article 51 of the Charter’.³⁷ This suggests, however, that, on the more fundamental preliminary point, the AG did not per se rule out that a Member State could be said to be implementing EU law in the sense of Article 51 of the Charter when acting on EU soft law.

14.2.4 Soft Law Interferences with Fundamental Rights in Practice

To make things more concrete, the present section will look into three distinct areas to illustrate how soft law may result in fundamental rights interferences. The first is the area of economic coordination and the Euro crisis response. To assist Eurozone Member States that are cut off from the international financial markets, the Eurozone Member States established the European Stability Mechanism (ESM) as an international organisation distinct from the EU. The ESM Treaty³⁸ confers a number of important tasks on the European Commission and the European Central Bank, which the CJEU accepted in the Pringle case.³⁹ The ESM funds made available to the Eurozone Member States are conditional on the negotiation and conclusion of a Memorandum of Understanding (MoU) between the ESM and the Member State concerned. These MoUs, while negotiated and signed by the Commission, follow from the ESM Treaty and therefore do not constitute challengeable acts under Article 263 TFEU.⁴⁰ At the same time, the EU can still be held liable for the Commission’s wrongful acts or omissions, specifically under the Charter (and the right to property, notably of those people affected by the austerity or restructuring measures taken in execution of the MoU), as confirmed by the Court in Ledra Advertising.⁴¹

The legal nature of the MoUs, negotiated by the Commission under the ESM, is highly disputed. The analogous MoUs agreed between the Commission and non-Eurozone Member States, under EU law, were considered to be ‘mandatory’ by the CJEU,⁴² against the findings of the AG,⁴³ but it is unclear whether this can be transposed to the ESM MoUs. AG Wathelet in Mallis suggested they are non-binding,⁴⁴ as does Repasi,⁴⁵ while Poulou suggests they are binding.⁴⁶ The Court in Ledra Advertising did not address this question and focuses on the Commission’s conduct in negotiating the MoU. Distinguishing the negotiation by the Commission from the MoU itself is in part built on a legal fiction however, and it seems difficult to see how the Commission’s conduct as such would constitute an interference in the absence of the MoU itself constituting such an interference. The Court of Justice indeed seems to assume that the MoU (as negotiated by the Commission) interfered with the right to property of the deposit holders at the banks that were put under resolution. Ledra would thus imply either that the CJEU accepts that a non-binding act may have such constraining practical or indirect legal effects that a fundamental rights interference may result from it or, alternatively, that it believed the MoU was binding. While Ledra concerned the right to property of deposit holders (of Cypriot banks), it is useful to flag that the MoU also prescribed an austerity programme with significant fundamental rights implications. For instance, the Cyprus MoU foresaw that Cyprus would reform its pension system, increasing the minimum age for pension entitlements and reducing pensions.⁴⁷

A second set of examples can be seen in the soft law adopted in the Area of Freedom, Security and Justice (AFSJ), for instance, by the EU Agency for Asylum (EUAA). Article 13 of its establishing Regulation provides that its Management Board can adopt operational standards, indicators, guidelines, and best practices to ensure a correct and effective implementation of Union law on asylum.⁴⁸ The agency thus adopts post-legislative guidance fleshing out different provisions of the EU asylum acquis, such as the Asylum Procedures Directive. In its Articles 14 and 15, the latter sets out the rule that applicants are entitled to a personal interview before a decision on their application is made as well as setting out the requirements that such a personal interview must meet.⁴⁹ Under Article 15(3)(c) of the Directive, the ‘communication shall take place in the language preferred by the applicant unless there is another language which he or she understands and in which he or she is able to communicate clearly’. In the EUAA’s Guidance on asylum procedure: operational standards and indicators of September 2019, however, Standard 30 prescribes that ‘the personal interview takes place in a language the applicant understands’.⁵⁰ The ‘good practice’ listed under that standard provides that if no interpreter is available in the language that the applicant understands, another language ‘that the applicant is reasonably expected to understand’ can be used. The guidance that is provided to national authorities to comply with the Directive thus effectively sets a lower (or at least less detailed) standard than the Directive itself. Since the latter is explicitly intended to promote inter alia the right to human dignity and the right to asylum,⁵¹ it cannot be excluded that the guidance might set the actual standard for the administrative practice of Member States and thereby interfere with these fundamental rights. This despite the Court in Addis ruling that the personal interview is of fundamental importance in the asylum procedure.⁵²

A further example in the AFSJ can be found in the EU’s response to Russia’s invasion of Ukraine. Quite quickly this led to calls within the EU of banning Russian tourists from acquiring Schengen visas, although the Schengen Visa Code⁵³ does not explicitly provide for such a ban. Following the EU Council’s suspension of the EU-Russia visa facilitation agreement,⁵⁴ Russian citizens are to be given the default treatment under the Visa Code regulation when applying for visas. When it comes to refusing Russian citizens’ applications, the main relevant provision in the Visa Code on which Member States would rely is Article 21(3)(d), which requires Member States to verify that ‘the applicant is not considered to be a threat to public policy, internal security or public … or to the international relations of any of the Member States’. While the regulation has not been amended, in September 2022 the European Commission did adopt guidance specifically for the issuance of visas to Russian citizens.⁵⁵ These guidelines seem to lower the bar for Member States to refuse applications from Russians: ‘As far as Russian nationals travelling for tourism are concerned, having a very strict approach is justified as it is more difficult to assess the justification for the journey.’⁵⁶ The Commission in its guidance also seems to lower the bar by introducing the notion of a ‘potential threat’, a concept that does not as such appear in the Visa Code or the (general) Visa Handbook.⁵⁷ While there is evidently no fundamental right of Russians to visit the EU, one may well argue that there is a fundamental right for Russian citizens, pursuant to Articles 20 and 21 of the Charter, not to be treated differently from, for example, Kazakh citizens (i.e., citizens of states other than Russia that are also listed in Annex I of Regulation 2018/1806⁵⁸ with which the EU has not concluded a facilitation agreement).⁵⁹ Arguably, the result of the guidance is that Member States are given a basis to routinely reject applications from Russians requesting a visa for touristic purposes.

A third illustration may be found in the digital sphere where the EU has adopted ambitious legislative packages in recent years. Although the General Data Protection Regulation⁶⁰ is not limited in scope to the processing of data through digital means, its importance for the digital information society is self-evident. In 2022, the EU legislator also adopted the Digital Services Act (DSA) and a Data Governance Act (DGA).⁶¹ Aside from their special relevance for the digital provision of services and the digital processing of data, two further features of these three different legislative instruments are noteworthy for the purposes of this chapter: they contain important enabling clauses foreseeing the adoption of soft law to ensure the proper application of the legislative acts and they rely on a new type of governance mechanism by setting up so-called Boards (the European Data Protection Board [EDPB], the European Board for Digital Services [EBDS], and the European Data Innovation Board [EDIB]) that are to adopt such soft law. While these Boards are heterogeneous in their nature, structure, and functioning,⁶² they have in common that they bring together representatives of the national authorities and that they are given the power to adopt soft law (guidance) on any issue coming under the scope of their respective legislative acts. This is very similar to the constellation at issue in the ThyssenKrupp cases discussed above. When it comes to the actual application of these legislative packages, different fundamental rights are evidently in play. The DSA is especially clear on this as it provides in its Article 34 that providers must make an assessment of how their services and the way they are offered may negatively affect rights, in particular the fundamental rights to human dignity enshrined in Article 1 of the Charter; to respect for private and family life enshrined in Article 7 of the Charter; to the protection of personal data enshrined in Article 8 of the Charter; to freedom of expression and information, including the freedom and pluralism of the media, enshrined in Article 11 of the Charter; to non-discrimination enshrined in Article 21 of the Charter; to respect for the rights of the child enshrined in Article 24 of the Charter; and to a high-level of consumer protection enshrined in Article 38 of the Charter.⁶³

How providers are concretely meant to mitigate risks is initially left to self-regulation, but Article 35(3) provides that the Commission ‘may issue guidelines … in relation to specific risks, in particular to present best practices and recommend possible measures, having due regard to the possible consequences of the measures on fundamental rights enshrined in the Charter of all parties involved’. In short, it may be expected that soft law will de facto determine the limits within which fundamental rights may be impacted by providers of very large online platforms and very large online search engines.⁶⁴ Generally, for these three acts, their actual implementation will be significantly steered by acts of soft law, adopted by either the Commission, the Boards, or the regulated entities themselves.

What the examples in these three very diverse policy fields show is, first, a seemingly insatiable appetite for soft law, not just on the part of the executive branch and, second, the very real possibilities for soft law to affect the legal position of natural and legal persons.

14.3 Remedies and Soft Law

The argument was made above that the scope of the Charter could, and in light of soft law’s practical effects should, extend to soft law. If that is indeed the case, the question becomes which remedies are available to challenge soft law’s interferences with fundamental rights. This section will highlight the two main existing judicial mechanisms that are available, as well as one administrative mechanism. The backdrop of this discussion is that soft law, by its nature, precludes the remedy of the action for annulment under Article 263 TFEU.

14.3.1 Unavailability of the Action for Annulment

The Court of Justice shut this door firmly in 2018 in Belgium v Commission,⁶⁵ despite suggestions by AG Bobek to take a more flexible approach. The Court of Justice thereby endorsed the findings of the General Court, which, at first instance, observed that the recommendation does not have and is not intended to have binding legal effects with the result that it cannot be classified as a challengeable act for the purposes of Article 263 TFEU.⁶⁶ The problematic repercussions of Belgium v Commission for effective judicial protection were decried by Arnull in strong terms: ‘[I]n direct actions the Court of Justice and the General Court now sometimes seem content to collude with other institutions to evade the requirements laid down by the Treaty.’⁶⁷ What therefore remains is the action for damages and the preliminary reference procedure, before the Courts, and the different forms of review by administrative bodies as a non-judicial remedy.

14.3.2 The Action for Damages

As the Ledra case discussed above suggests (assuming the CJEU found the MoU to constitute soft law), it is in principle possible to challenge the fundamental rights interferences of soft law through an action for damages. It is established case law of the EU Courts to require three cumulative conditions for the EU to incur non-contractual liability, those being the unlawfulness of the EU’s conduct, the fact of damage, and the existence of a causal link between the conduct and the damage.⁶⁸ In Ledra, the Court dismissed the claim for damages on the first of these conditions by finding that the Commission had not committed a sufficiently serious violation of the fundamental right involved because the restriction of the fundamental right could be justified pursuant to Article 52(1) of the Charter.⁶⁹ Specifically for soft law, the third cumulative condition appears to be the most problematic, however. Under the Court’s established case law, a sufficiently direct causal link means that the conduct must be the determining cause of the damage.⁷⁰ Yet, given its non-binding nature, the indirect legal and practical effects of soft law seem incapable of ever constituting such a determining cause.⁷¹ This is clear where EU soft law is subsequently implemented by further decisions (binding or non-binding) but also when it results in the first and fourth scenarios noted above in Section 14.2.3. The threshold to show that soft law itself determined a private party’s behaviour or the factual conduct of a government body lies exceptionally high. Where the action for annulment against soft law will be inadmissible, an action for damages might be admissible but would then always fail on the merits.

14.3.3 The Preliminary Reference Procedure

The only judicial remedy practically available will then be the preliminary reference procedure, as evidenced by the FBF case.⁷² In that case, the Court followed up on Belgium v Commission mentioned earlier where it refused to review soft law in an action for annulment but where it noted that ‘Article 267 TFEU confers on the Court jurisdiction to deliver a preliminary ruling on the validity and interpretation of all acts of the EU institutions without exception’.⁷³ From a more institutional perspective, the Court’s decision, as also predicted by AG Bobek in his Opinion in FBF,⁷⁴ goes against the current evolution of the judicial system at EU level. In terms of workload, it would have made more sense to allow for a direct review of soft law measures, given the recent expansion of the General Court. That the Court confirmed that there would be a possibility to seize it in order to challenge the validity of soft law was still welcomed by Gündel,⁷⁵ but there are at least three important constraints, inherent to the preliminary reference procedure, that require pointing out. First, the preliminary reference procedure requires a reference point in the national legal order that can be relied on to seize a national judge; second, a procedure must also be available at national level (which may not be self-evident in the realm of soft law); and third, the preliminary reference procedure is not a self-standing remedy but instead depends on the national judge referring questions.

The first two of these constraints are intertwined: concretely, a natural or legal person taking issue with an EU soft law measure will only be able to indirectly challenge this measure before a national judge if there is a national measure or conduct that is challengeable pursuant to a procedure available in national law. In the first scenario noted above in Section 14.2.3, such a reference point may not be available. This puts an acute challenge to the right of natural and legal persons to effective legal protection. The avenue to which the Court directs applicants revives the same problems that resulted from Plaumann and the Court’s response in cases like UPA and Jégo Quéré,⁷⁶ as well as revealing a further challenge to the principle of procedural autonomy.

In the scenario where EU soft law will be followed up on or implemented at national level through further (national) soft law, this may prove problematic. While different national legal systems seem increasingly open to allow such challenges,⁷⁷ this option still does not seem to exist in the majority of national legal systems. In addition, the requirements to be fulfilled by applicants may vary greatly between Member States. Should these differences simply be accepted in light of national procedural autonomy? That is doubtful. While national procedural autonomy should be considered a legal principle, rather than a temporary state of affairs in EU law, it is still to be balanced with other principles such as that to effective legal protection.⁷⁸ In this regard, Arnull has noted that in the post-Lisbon era ‘the venerable principles of national procedural autonomy, equivalence and effectiveness seem to have been absorbed into a more complex matrix of rules and principles which represent a considerable intrusion into fields formerly considered the prerogative of the Member States’.⁷⁹ In line with UPA, Unibet,⁸⁰ and Article 19(1) TEU,⁸¹ it would instead be up to the Member States to provide adequate remedies. Assuming that the Court will refuse to revisit its Belgium v Commission ruling, it will be up to the Member States courts to fix any resulting lacunae. This is already a significant requirement imposed on those legal systems that have not (yet) accepted the reviewability of national soft law endogenously. However, it will be most acute in those cases under a second scenario, in the vein of Jégo-Quéré, where there is no national measure (not even a soft law measure) to be contested to begin with.

The third constraint of the Court’s solution in FBF is that, under the established case law of the Court of Justice,⁸² the preliminary reference procedure is not a self-standing remedy offered to (private) parties. It is rather an instrument through which national courts may enter into dialogue with the Court of Justice and whether preliminary references are sent to the Court ultimately depends on those national judges, not on the parties appearing before them. FBF therefore means that a genuine remedy against soft law will depend on national judges referring questions on validity to the Court of Justice.

14.3.4 Subjecting Soft Law to Administrative Review

In light of the constraints that come with the preliminary reference procedure identified in Section 14.3.3, it is important to draw attention to other possible remedies. As Jääskinen observed, there are indeed ‘alternative ways of constitutionally legitimised control structures of the use of public powers’ and ‘judicial protection is a societally and economically scarce resource which cannot be light-heartedly allocated to cases that are better dealt with [by] other types of remedies or that do not deserve the attention of courts’.⁸³ While the possibility to seize an independent judicial tribunal would appear to be a requirement under Article 47 of the Charter, there may indeed be other (administrative) remedies available that may be more effective. Apart from the general possibility to seize the Ombudsman (when a soft law act is alleged to be vitiated by an instance of maladministration), such administrative remedies arguably ought to bear the brunt of ensuring adequate legal protection, with the judicial remedy acting as a safety valve.

The EU legislator has in the past already experimented with administrative review procedures that potentially allow soft law to be challenged. The typical constellation in which such review procedures have been established is one where an EU agency has been granted certain powers and where the Commission has been identified as the body competent to review its acts at the request of a private party. Pre-Lisbon, such clauses were included to ensure the possibility of at least some review of agency decisions, given the agencies’ unclear passive legal standing in the action for annulment.⁸⁴ Post-Lisbon, such clauses are not strictly needed anymore, which explains why they were deleted in the 2019 revisions of the Eurofound, Cedefop, and EU-OSHA Regulations and in the 2017 European Union Intellectual Property Office Regulation.⁸⁵ Remarkably, the revision in 2022 of the establishing regulation of the European Centre for Disease Prevention and Control (ECDC) revamped this procedure, rather than deleting it from the regulation all together.⁸⁶ Article 28 of the ECDC Regulation now provides:

Article 28

Examination of legality

1. 1. Member States, members of the Management Board and third parties directly and individually concerned may refer any act of the Centre, whether express or implied, to the Commission for examination of the legality of that act (‘administrative appeal’).

2. 2. Any administrative appeal shall be made to the Commission within 15 days of the day on which the party concerned first became aware of the act in question.

3. 3. The Commission shall take a decision within one month. If no decision has been taken within that period, the administrative appeal shall be deemed to have been dismissed.

4. 4. An action for annulment of the Commission’s explicit or implicit decision referred to in paragraph 3 of this Article to dismiss the administrative appeal may be brought before the Court of Justice of the European Union in accordance with Article 263 TFEU.

Similar mechanisms are still in place for the Community Plant Variety Office (CPVO)⁸⁷ and for the European Food Safety Authority (EFSA).⁸⁸ While part of the administrative appeal is modelled on Article 263 TFEU, notably the requirement that private parties need to be directly and individually concerned, the scope of the appeal seems to be broader, since any act of the agency may be referred to the Commission.⁸⁹

In 2019, the EU legislator further experimented with this type of review mechanism by introducing Article 60a in the three Regulations establishing the European Supervisory Authorities (ESAs) in the financial sector.⁹⁰ The new provision was introduced specifically to allow for a review of some of the soft law that the ESAs adopt.⁹¹ Article 60a provides: ‘Any natural or legal person may send reasoned advice to the Commission if that person is of the opinion that the Authority has exceeded its competence, including by failing to respect the principle of proportionality referred to in Article 1(5), when acting under Articles 16 and 16b, and that is of direct and individual concern to that person.’⁹² Although it has already been in force for a couple of years, this administrative review procedure has not actually been relied on yet.

For a number of reasons, it is also doubtful whether that procedure offers any genuine remedy against the ESAs exceeding their competences through the adoption of soft law. First, only individually and directly concerned persons may communicate reasoned advice to the Commission per Article 60a. This arguably makes the remedy dependent on similar standing requirements as those under Article 263 TFEU and the earlier administrative review procedures, but it is unclear how these would apply in the given context, as soft law acts cannot in principle be of direct concern to a person. After all, in the Courts’ established case law, direct concern means that a measure ‘must directly affect the legal situation of the individual and, second, it must leave no discretion to its addressees who are entrusted with the task of implementing it, such implementation being purely automatic and resulting from the EU rules alone without the application of other intermediate rules’.⁹³ Since the effects of soft law are merely of a practical nature or, at the most, indirect legal effects, it seems a priori impossible for soft law to ever be of ‘direct concern’. In addition, if the Plaumann criteria would also apply mutatis mutandis, it would be almost impossible for a party to demonstrate that they are individually concerned by the ESAs’ guidelines and Q&As since these will be general in scope.⁹⁴ This may be different for the ESAs’ recommendations, as Article 16 of the ESA Regulations provides that they may be addressed ‘to one or more [national] competent authorities or to one or more financial institutions’. Still, since direct and individual concern are cumulative conditions, the standing requirement will never be met for soft law. Unless, of course, ‘direct and individual concern’ is not to be understood in the same way as the analogous notions in Article 263 TFEU.

Second, Article 60a does not provide an actual remedy. Even if an admissible complaint were to be lodged, the Commission cannot, under Article 60a or any other provision of the ESAs Regulations, provide an actual remedy where it finds that an ESA has exceeded its competences. Only ESAs can issue and, in turn, repeal guidelines, recommendations, and Q&As and the Commission cannot give instructions to the ESAs or override the soft law, except where it has been specifically empowered to adopt delegated or implementing acts. Article 60a therefore at best represents an alert mechanism. While the attempt of the legislator to afford private parties protection against the effects of the ESAs’ soft law should be lauded, the effectiveness of Article 60a of the ESAs Regulations will be undermined by the legislator’s failure to recognise that the challenges posed by soft law represent a paradigm shift for the system of remedies. Transplanting admissibility requirements from procedures that aim at the review of hard law ignores the fundamental features characterising soft law. While administrative review of soft law seems an appropriate way forward (see below), it would have to be devised mindful of the specific features of soft law.

14.3.5 A Possible Way Forward

Building on the legislator’s recognition, in the 2019 ESAs Regulations, of the importance of a possibility to review soft law, complementing the existing EU system of remedies with a dedicated administrative review mechanism appears to be the way forward. Special attention should thereby be devoted to three issues: access to the mechanism, the nature of the adjudicating body, and the remedy provided.

As noted in the discussion of Article 60a of the ESAs Regulations, the mechanism’s admissibility requirements should take proper account of soft law’s non-binding nature and therefore not replicate those applicable under Article 263 TFEU. Instead, the mechanism could be open to parties showing an interest.

As regards the body made responsible to review soft law, the 2019 revision of the ESAs Regulations also provides interesting clues. During the legislative negotiations, the European Parliament had proposed to extend the mandate of the ESAs’ Joint Board of Appeal, making it competent to hear challenges to the ESAs’ soft law.⁹⁵ The Parliament’s suggestion did not make it in the end, but it is interesting, nonetheless. Within decision-making EU agencies, there are already specialised and independent boards of appeal that are competent to review (individual) binding decisions.⁹⁶ The 2019 revision thus constituted a missed opportunity to tap this potential and it would be worthwhile to revisit this possibility. For those EU agencies adopting soft law and who are already equipped with a Board of Appeal, the latter’s mandate could be broadened.⁹⁷ Where soft law is adopted by other agencies or bodies of the EU or by the Commission itself, a variant of Article 60a of the ESAs Regulations could be envisaged where the review is entrusted to a functionally independent entity within the Commission. Just like the independent Regulatory Scrutiny Board (unilaterally established by the Commission) double-checks the soundness of the Commission’s proposals and impact assessment, so could an independent review board be entrusted with scrutinising soft law.⁹⁸

Lastly, any mechanism should provide for a clear remedy. Given soft law’s non-binding nature, that remedy need not mimic the remedy under Article 263 TFEU, just like the admissibility requirements need not be mimicked. Instead, a requirement for the original author to reconsider the soft law in light of the reviewing authority’s remarks could also be sufficient.

14.4 Conclusion

The present chapter looked into how the informalisation of governance through the adoption of soft law can affect the fundamental rights position of private parties. After briefly exploring the nature and function of soft law, the primordial question for this chapter to address was how soft law, given its formally non-constraining nature, is relevant from a fundamental rights perspective.

Although the jurisprudence of the ECtHR and the CJEU is not explicit on this point, the chapter argued that the possibility of fundamental rights interferences by soft law (and concomitantly the need for review) should be accepted by drawing an analogy to the CJEU’s case law on the fundamental freedoms in the internal market. To further bring this point home, four examples from three very different policy areas were presented to illustrate possible soft law interferences with fundamental rights.

Subsequently, the chapter looked into the remedies available in the EU in relation to soft law, starting with the judicial remedies. Since the latter are premised on the idea that government acts through binding measures, they do not cater to the review of soft law. Only with the recent FBF case has the Court of Justice left open one avenue to assess the legality of soft law, albeit that the preliminary reference procedure on its own cannot secure a watertight system. As a result, a further fundamental rights interference could result, since the right to an effective remedy may not be guaranteed for all instances where soft law is adopted at EU level.

The chapter then looked into the possibility of relying on non-judicial remedies for challenging soft law, highlighting how the EU legislator has already experimented with this approach. While the legislator’s attempts do not seem to have been fully thought through, they should be supported and built upon, since extra-judicial administrative review of soft law seems more attuned to challenging soft law than going through the judicial avenue of the preliminary reference procedure.