1.Introduction

When EU law rules are enforced by multiple authorities in a multi-layered administrative system, a key element in ensuring enforcement effectiveness is the concept of trust. For example, as will become apparent, the concept of trust is revealed at different stages of the enforcement process: the public enforcement system of the EU antitrust prohibitions (Articles 101 and 102 TFEU, relating to cartels and other anti-competitive agreements and abuse of dominance) entails concurrent enforcement action by both the Commission and the national competition authorities (hereinafter NCAs). In such a setting, trust may be established when an authority can comfortably regard itself as relieved from acting, while knowing that a given case can be adequately dealt with by another authority, based on the same TFEU provisions. From a different standpoint, when parallel enforcement action is undertaken by multiple authorities, the concept of trust is again important; these provisions must be applied uniformly by these entities. Furthermore, trust is important in this context for the cooperation between these enforcers, which unfolds at different stages of the multi-level enforcement process. Consequently, solid mechanisms which allow authorities to support each other’s enforcement actions are essential.

The flexible workload division system, in which enforcement is shared between the various EU and national authorities, is a good ‘setting’ in which mutual trust may be studied, especially given the recent trend of pushing more enforcement work from the centre to the domestic authorities.Commission, ‘COMMUNICATION FROM THE COMMISSION TO THE EUROPEAN PARLIAMENT AND THE COUNCIL – Ten Years of Antitrust Enforcement under Regulation 1/2003: Achievements and Future Perspectives’ (Communication) COM (2014) 453 final, ch 2, para 8. This trend, which highlights the dynamics of the multi-layered enforcement system, may be progressively observed in various landmarks that contribute to developing this idea of trust: for example, the decentralisation brought about by Regulation 1/2003,Council Regulation (EC) 1/2003 of 16 December 2002 on the implementation of the rules on competition laid down in Articles 81 and 82 of the Treaty [2002] OJ L1/1. the soft-law actions meant to smoothen the Commission and NCAs relationship, the EU Courts’ case-law, arguably furthering the trust between these actors.

The recent (ECN+) Directive 2019/1Directive (EU) 2019/1 of the European Parliament and the Council of 11 December 2018 to empower the competition authorities of the Member States to be more effective enforcers and to ensure the proper functioning of the internal market [2019] OJ L11/3 (Directive). is meant to render NCAs as more effective enforcers of the EU antitrust prohibitions. In this contribution we aim to determine whether the Directive furthers the trust already established through the hard-, soft-, and case-law initiatives that have been mentioned above. To this end, this paper tracks the development of the trust in the NCAs’ (EU) antitrust enforcement work and assesses the building-blocks on which trust is shaped. It then evaluates the Directive’s core elements, which are prone to increase trust, in order to test whether the Directive embodies a correct follow-through of the hard-, soft-, and case-law. This paper also looks into whether serious enforcement gaps still remain, gaps which undermine the trust between the European antitrust enforcers, and consequently the Directive’s core objective of rendering NCAs effective enforcers.

In this paper, the concept of trust is key to explaining the Commission and NCAs relationship as far as the enforcement of the EU antitrust prohibitions is concerned. This trust relationship was built in the past fifteen years of enforcement experience in the EU, on the basis of constant dialogue between the enforcers, legislators, and stakeholders. As this paper will reveal, this dialogue is present at different stages. Firstly, when designing and / or amending the enforcement regime(s), and secondly, at various steps of the actual enforcement process. Regarding the former, dialogue is at the core of the provisions of Regulation 1/2003. One may think of the respect paid to the Member States’ institutional autonomy which was meant to preserve national specificities in the enforcement exercise. Furthermore, dialogue lied also at the heart of the public consultation, which shaped the adoption of Directive 2019/1, essentially an instrument, which answered the NCAs’ call for a minimum EU law-based framework for effectively carrying out their tasks.‘Summary report of the replies to the Commission’s Public Consultation on Empowering the national competition authorities to be more effective enforcers’ (2015) 4, http://ec.europa.eu/competition/consultations/2015_effective_enforcers/Summary_report_of_replies.pdf, accessed 29 May 2019. Regarding the latter, namely the actual enforcement process, dialogue is exhibited in the context of case allocation, exchange of information, performing of fact-finding and investigation measures, imposing sanctions, etc.

This framework allows one to establish which factors or indicators are prone to further increase enforcement trust. However, trust is not something that can be easily measured in numbers. Trust is not static data and its presence or absence may be deduced from a multitude of indicators; trust must be viewed in the correct context. For example, a high number of enforcement decisions does not necessarily say anything about how an investigation was performed, and if the NCA had all the tools they needed at their disposal for the enforcement exercise. Instead, trust may be reflected in the powers that an authority is given, and in the autonomy that is afforded to that authority. At the same time, unlimited powers do not automatically mean unlimited trust. Accountability requirements are also important when multiple authorities are involved in applying the same rules, as it is important to ensure the credibility of their actions and to protect against arbitrariness. Such indicators of trust will be further discussed in the following paragraphs.

2.Multi-level administration in EU antitrust public enforcement

The concept of multi-level administration is used to describe the interplay between different actors (Commission, NCAs, national courts)In this paper we focus mainly on the administrative enforcement aspects. involved in the public enforcement of the EU antitrust provisions, across different levels (European, domestic), and through horizontal and vertical relationships between those actors.See for a similar definition, Antonio Manganelli, Antonio Nicita, Maria Allessandra Rossi, ‘The Institutional Design of European Competition Policy’ (2010) EUI Working Papers RSCAS 2010/79, https://fsr.eui.eu/Documents/WorkingPapers/ComsnMedia/2010/WP201079.pdf, accessed 7 June 2019. As detailed below, the rules governing these relationships are embedded in hard- and soft-law instruments. Analysing these instruments and the ensuing dynamics of the multi-layered enforcement system will contribute to clarifying the relationship between the different enforcement entities. This is important for effective enforcement because the authorities cooperate in many areas, ranging from case allocation to mutual enforcement assistance. Furthermore, the analysis will allow to identify the areas and means by which the multi-layered administrative enforcement of EU antitrust law may be improved.

Articles 101 and 102 TFEU (initially Articles 85 and 86 TEEC, and subsequently Articles 81 and 82 TEC) have been in force since 1958. Already then, the Commission was granted a central role in the application of the principles laid down in these Treaty provisions.EC Treaty (Treaty of Rome, as amended) art 89. This general supervisory role of the Commission was also recognised by the General Court in Case T-24/90 Automec EU:T:1992:97, para 74. Regulation 17/62Council Regulation (EEC) 17 implementing Articles 85 and 86 of the Treaty [1962] OJ 13/87. was adopted to ensure the enforcement of these provisions. According to this Regulation, NCAs remained competent to apply Articles 85(1) and Article 86 TEEC alongside the Commission;Ibid [9(3)]. however, the Commission was granted the exclusive power to issue exemption decisions pursuant to Article 85(3) TEEC.Ibid [9]. The EU has grown considerably since the time of the adoption of Regulation 17. Therefore, in the early years, it could have been more appropriate to have a strong central authority enforcing the antitrust rules.Silke Brammer, Horizontal aspects of the decentralisation of EU competition law enforcement (Leuven 2008) 8. However, with the enlargement of the EU, the complexity of the legal and economic environment in the EU transformed, given the increase in diversity of legal regimes of the Member States. The question arose whether the central enforcement system was still suitable in this new environment.See also Clause-Dieter Ehlermann, ‘Implementation of EC competition law by national anti-trust authorities’ [1996] 17 ECLR 2, 88-95. The main criticism the Commission received regarding Regulation 17 concerned the high backlog of cases, the length of procedures, and the lack of transparency and motivation of comfort letters (i.e. the means through which the Commission was informing the undertakings that it found no grounds for action based on the EU antitrust provisions).Brammer (n. 11); Frank Montag, ‘The case for a reform of Regulation 17/62: Problems and possible solutions from a practitioner’s point of view’ [1998] 22 Fordham International Law Journal 3, 826; Alison Jones & Brenda Sufrin, EU competition law. Text, cases, and materials (Oxford University Press 2016) 887; S. Macro Colino, Competition law of the EU and the UK (Oxford University Press 2011) 74; Giorgio Monti, EC competition law (Cambridge University Press 2007) 395 et. seq.; Clause-Dieter Ehlermann, Laraine Laudati (ed.), European competition law annual 1997 – Objectives of competition policy (Hart Publishing 1998) 567 et. seq. The failed attemptsE.g. Commission, Commission Notice on cooperation between national courts and the Commission in applying Articles 85 and 86 of the EC Treaty [1993] OJ C39/6; Commission Notice on cooperation between national competition authorities and the Commission in handling cases falling within the scope of Articles 85 and 86 of the EC Treaty [1997] OJ C313/3. to address these issues only increased the pressure to move away from the centralised enforcement system.Brammer (n. 11).

Regulation 1/2003 arguably represents a turning point in the modernisation of the EU antitrust enforcement, because it created a system where the Commission, the Member States’ administrative and judicial bodies together enforce the material EU antitrust rules. The centralised individual exemption regime of Article 101(3) TFEU was abolished, and NCAs and national courts are now able to apply Article 101 TFEU in its entirety. When NCAs apply their own national competition rules, they must apply the EU provisions in parallel, when the trade between the Member States is impacted.Council Regulation (EC) 1/2003 of 16 December 2002 on the implementation of the rules on the competition laid down in Articles 81 and 82 of the Treaty [2002] OJ L1/1, art 3(1). This setting yielded important consequences for both the Commission and the NCA:For thorough discussions on the impact of Regulation 1/2003, see Wouter Wils, ‘Ten years of Regulation 1/2003 – A retrospective’ [2013] 4 Journal of Competition Law and Practice 4; Katalin Cseres, ‘Multi-jurisdictional competition law enforcement: The interface between European competition law and the competition laws of the new Member States’ [2007] 3 European Competition Journal 2; Francisco Marcos, Albert Sanchez Graells, ‘A missing step in the modernisation stairway of EU competition law – Any role for block exemption regulations in the realm of Regulation 1/2003?’ [2010] 6 The Competition Law Review 2, 188-189 and the cited literature. for example, it allowed more prioritisation room for the Commission and more efficient use of the enforcement resources, since it was generally perceived that action by multiple enforcers may be more effective, especially given the domestic authorities’ knowledge of the domestic markets.Catalin Stefan Rusu, ‘The real challenge of boosting the EU competition law enforcement powers of NCAs: in need of a reframed formula?’ [2018] 13 The Competition Law Review 1, 30. Thus, the decentralisation exercise also resulted in building a strong conviction that NCAs become ‘EU agencies’, rather than simply national enforcement bodies, when they enforce EU law.See Rusu (n. 18) with reference to N. Fennelly, ‘The national judge as judge of the European Union’, in: The Court of Justice and the Construction of Europe: Analyses and Perspectives on Sixty Years of Case-Law (Springer 2013) and Giorgio Monti, ‘A plea for ‘extraterritorial’ antitrust enforcement by National Competition Authorities’ [2014] Keynote speech at the Fourth ACELG Annual Conference.

Nevertheless, the challenge of a multi-level enforcement system guarantees that each authority enforces the substantive rules uniformly, ensuring a level playing field, despite the involvement of different actors. The administrative authorities belong to different jurisdictions and have different competences embedded in their national regimes. Naturally, legal certainty for companies active in the Internal Market could be at stake. Furthermore, it is important that NCAs have the same guarantees of independence and enforcement tools when enforcing antitrust infringements under national or EU competition law. Otherwise, the parallel application of these provisions could lead to different outcomes in the same, or similar cases.

The system created by Regulation 1/2003 relies heavily on the Member States’ domestic procedures, used by NCAs when enforcing the EU antitrust provisions. This means that important matters (discussed below) are left to the institutional and procedural autonomy of the Member States, subject to the EU principles of equivalence and effectiveness:Case 33/76 Rewe EU:C:1976:188, para 5 and Case 45/76 Comet EU:C:1976:191, para 13; see also Case 432/05 Unibet EU:C:2007:163, para 39 and Case C-40/08 Asturcom EU:C:2009:615, para 41. i.e. domestic procedural rules may not discriminate against, or render practically impossible or excessively difficult to pursue, EU law-based actions.Case C-261/95 Palmisani EU:C:1997:351, para 27. However, regardless of these principles, the fact that Regulation 1/2003 does not provide an institutional and procedural framework for the functioning of NCAs, a matter which is left mostly to domestic law, leads to diverging regimes across the Member States. This may create obstacles to the effective and uniform enforcement of EU competition law. Such shortcomings are clearly spelled out in Recitals 5 to 7 of Directive 2019/1.

Furthermore, because the authorities enforce the same substantive rules, cooperation and fostering information exchanges between the various authorities is an important tool to ensure the uniform application of Articles 101 and 102 TFEU. Regulation 1/2003 lays down rules in this respect.Council Regulation (EC) 1/2003 of 16 December 2002 on the implementation of the rules on the competition laid down in Articles 81 and 82 of the Treaty [2002] OJ L1/1, ch 16. To further ensure enforcement coherency and uniformity and to allow stakeholders to benefit from a more level playing field, the European Competition Network (ECN) was created.Commission Notice on cooperation within the Network of Competition Authorities [2004]

OJ C101/43 (Network Notice). The ECN provides the framework for cooperation between the NCAs and the Commission by stimulating communication and exchange of perspectives. This brings to light the dynamics of the multi-layered enforcement system, while also fostering the creation of a common competition culture in Europe, despite the diversity that characterises the ECN, which is essentially a forum composed of multiple enforcers, with different structures and approaches to the enforcement process.Ibid [1]-[2]. Nevertheless, the Network Notice iterates that the NCAs have recognised the standards of each other's systems as a basis for cooperation, which should not be perceived as an end in itself, but rather as an instrument to achieve efficient and flexible work division and an effective and consistent application of the EU antitrust rules.See also Commission Notice on the handling of complaints by the Commission under Articles 81 and 82 of the EC Treaty [2004] OJ C101/65, para 20. In this respect, the Notice pushed towards some voluntary harmonisation of certain aspects of national procedural regimes.Firat Cengiz, ‘Multi-level governance in competition policy: the European Competition Network’ [2010] 35 European Law Review 5, 669; Giorgio Monti, ‘Strengthening national competition authorities’ [2018] 13 The Competition Law Review 2, 103. However, the ECN is based on non-binding soft-law rules, the application of which is often non-transparent to the outside world, again placing legal certainty in a vulnerable position.This matter has been signalled during the 2015/2016 public consultation ‘Empowering the national competition authorities to be more effective enforcers’, http://ec.europa.eu/competition/consultations/2015_effective_enforcers/index_en.html, accessed 20 May 2019. For example, the reply of the Irish Competition and Consumer Protection Commission, 9, states in no unclear terms: ‘The CCPC considers that legislative action at EU level is necessary to empower NCAs to be more effective enforcers, as non-legislative soft-law measures that have been implemented to date have not been effective in ensuring consistency of approach in relation to enforcement of competition law in Member States across the EU.’ As section 3.2 of this contribution will further detail, legal certainty is important for the market players too, especially when it comes to knowing which authority will investigate their conduct, or to which authority they can submit a complaint. The discussion below will highlight (some of) the shortcomings of using soft-law criteria in case allocation. For a thorough analysis of the impact of Commission-issued soft-law on the domestic enforcement of EU competition law, especially from a consistency and legal certainty standpoint, see Zlatina R. Georgieva, Soft law in competition law enforcement and its reception in member states’ courts (Tilburg 2017), https://pure.uvt.nl/ws/portalfiles/

portal/23090447/Georgieva_Soft_law_28_06_2017_emb_tot_31_1_2018.pdf://core.ac.uk/download/pdf/34435674.pdf, accessed 17 May 2019. Due to the fact that the multi-level administration relies greatly on the NCAs’ work and because there is likely untapped potential for the NCAs to become more effective enforcers, the Commission presented the ECN+ Directive Proposal.Proposal for a Directive COM (2017) 142 final to empower the competition authorities of the Member States to be more effective enforcers and to ensure the proper functioning of the internal market [2017].

3.Building trust in the enforcement of the EU antitrust rules

3.1.Regulation 1/2003, decentralisation and soft-law action

Despite bringing about an important enforcement overhaul, Regulation 1/2003, was not always praised as the revolutionary piece of legislation it aimed to be, because, for example, the decisional output of the Commission has only marginally changed when comparing the numbers under Regulation 17/62 and under Regulation 1/2003.Wils (n. 17) 15, 16 shows that whereas under Regulation 17, the Commission adopted on average 7.5 prohibition decisions and 2 exemption decisions with conditions or obligations per year, it has under Regulation 1/2003 adopted on average 7 prohibition decisions and 3 commitment decisions per year. See also Monti (n. 19). Nevertheless, Regulation 1/2003 added value for the enforcement trust relationship built in the multi-layered enforcement system.

First, the joint enforcement of Articles 101 and 102 TFEU is a development that should not be taken lightly. Especially in the early 2000s, it was not that common in other areas of law that wide decision-making powers relating to EU law were entrusted with national authorities. Article 5 of the Regulation empowered NCAs specifically to adopt decisions finding infringements, ordering interim measures, accepting commitments, and imposing fines and penalties. Also, Article 35 left at the Member States discretion to designate their (administrative or judicial) NCAs, with the only EU law safety-net in this respect being that Regulation 1/2003 is effectively complied with. In our view, such bold institutional and enforcement-related assigning of powers signals the EU’s trust in the national entities’ capability of enforcing the EU rules. However, this trust is not unlimited since NCAs have no power to adopt negative decisions based on the EU antitrust provisions. The European Court of Justice (ECJ) clarified in the 2011 Tele2 Polska preliminary rulingCase C-375/09 Tele2 Polska EU:C:2011:270. that this power rests only with the Commission, based on Article 10 of Regulation 1/2003. Thus, the ECJ drew the boundaries of how far the decision-making trust can span, by choosing to safeguard against the risks of non-uniform application of EU antitrust law. In the same vein, the provisions of Article 16 of Regulation 1/2003 may be mentioned: NCAs cannot take decisions running counter to the decision adopted or contemplated by the Commission in the same proceedings.This provision essentially codifies the Case C-344/98 Masterfoods EU:C:2000:689.

However, secondly, trust also perspires from the more technical enforcement provisions of Regulation 1/2003, such as Chapter IV on the Commission and NCAs cooperation, while applying Article 101 and 102 TFEU (e.g. Article 11(1)). Such provisions lay the ground rules of the institutional and enforcement dynamics unfolding in the multi-layered enforcement system. For example, one may mention the multi-directional (Commission – NCAs, NCA – NCA) exchange of documents and information relating to the start, the development, and the conclusion of EU and domestic proceedings (Articles 11(2), 12(1)), and the possibility of consulting the Commission on any case involving the application of EU law (Article 11(5)). These tools arguably build trust between the concerned enforcers, as the Commission and the NCAs will only engage in such exchanges of information if they are confident that the information provided will be treated with the observance of safeguards in place for securing effective enforcement. At the same time, these cooperation tools enhance transparency in the multi-layered enforcement system. Yet, once again, this trust is again not unlimited: the initiation of Commission proceedings relieves the NCAs of their competence to apply Articles 101 and 102 TFEU (Article 11(6)). Although never officially used,Monti (n. 26) 103. this power essentially affords the Commission all the means to act, when it wants to do so.

Thirdly, a great degree of trust stems also from the Regulation’s provisions allowing competition authorities to help each other in performing investigations and fact-finding actions: NCAs consult on and assist the Commission’s so-called ‘dawn raids’ (Articles 20, 21), and may perform inspections in their territory and while using their domestic laws and procedures, on behalf of other enforcers (Article 22). This scenario specifically highlights a high degree of trust between these authorities, because such investigations are performed on the basis of domestic laws ‘foreign to’ the requesting authority, obviously under the safety-net of the effectiveness and equivalence principles.

Consequently, Regulation 1/2003 ensured a good breeding ground for enforcement trust to flourish. Yet, its trust-enhancing provisions, far-reaching as they may seem, were still not cohesive enough.See for example the explanations provided in the Network Notice (n. 23) [2] where the differences between various domestic systems are acknowledged. This is why the Commission felt the need to further elaborate on some enforcement aspects through soft-law,See http://ec.europa.eu/competition/antitrust/legislation/legislation.html, accessed

12 December 2018, for a comprehensive overview of the Commission’s soft-law mechanisms adopted since 2004. which set a clearer tone for the enforcement mutual trust than Regulation 1/2003 did. For example, the Network Notice builds on the trust-enhancing (technical) provisions of Regulation 1/2003, by providing handy practical examples of, and guidance on: information exchanges between enforcers, the use of confidential information, assisting investigations of other authorities, the position of complainants and of the undertakings investigated, etc. Remarkably, the Notice deals extensively with the principles and mechanisms of case allocation. This makes sense, because the cooperation during antitrust investigations unfolds only once cases are ‘properly’ allocated to one or more authorities. We will discuss the relevant aspects in this respect, in the following section.

3.2.Trust, case allocation, and the case-law of the Union Courts

In order to highlight how the Commission and NCAs trust relationship developed in practice, one may, for example, investigate when and how the Commission can reject handling complaints, thus leaving the enforcement task in the hands of the NCA(s). The Network Notice explains why such a flexible approach is important for competition authorities: they may reject cases for different reasons, they may want to reopen suspended proceedings at a later stage, or to close proceedings and transfer information to other authorities.Network Notice (n. 23) [22]. As already provided above, essentially, trust may be established when authorities can comfortably regard themselves relieved from acting, while knowing that other authorities can adequately handle the enforcement tasks. Next, we will look at under which circumstances can this be done in practice.

First, the case allocation process provides avenues to this end. The Network Notice (para 5) highlights the parallel enforcement competences and the shared responsibility for efficient work division between the competition authorities in the EU, whereas each authority retains full discretion in deciding whether or not to investigate a case. Yet, there are no hard-and-fast rules regulating work sharing,Commission Notice (n. 25).  since an authority may choose to act only when it is well-placed to deal with a case (i.e. when material links exist between the alleged infringement and a specific territory).Network Notice (n. 23) [8], [9]; Commission Notice (n. 25) [22]. The Commission, which is essentially well-placed to act if effects on competition may occur in more than three Member States, may nevertheless exercise its discretion not to act, thus leaving the case in the hands of the NCAs, either when NCAs are willing and ready to step in, or based on the so-called Automec EU interest formula, which requires a cumbersome balancing act embedded in an adequately reasoned decision:Case C-119/97 P UFEX v Commission EU:C:1999:116, para 89 et. seq. the Commission must weigh the significance of the alleged infringement regarding the functioning of the Internal Market, against the probability of establishing an infringement, and the extent of the necessary investigative measures. Thus, the Commission has plenty of room to trust NCAs in handling antitrust cases based on Articles 101 and 102 TFEU. And it has done so in the past, since after 2004 the NCAs adopted more than 85% of the antitrust decisions dealing with these TFEU articles.

Second, the relationship of trust in allocating cases between the Commission and the NCAs is built also on Article 13 and Recital 18 of Regulation 1/2003, which allow NCAs and the Commission to refuse enforcement action if the case is or has been dealt with by another authority. The Si.mobil, easyJet, Trajektna luka Split and Agria Polska judgments of the General Court (GC) and the ECJ’s appeal ruling in Agria PolskaCases T-201/11 Si.mobil v Commission EU:T:2014:1096; T-355/13 easyJet Airline v Commission EU:T:2015:36; T-70/15 Trajektna luka Split d.d. v Commission EU:T:2016:592; T-480/15 and C-373/17 P Agria Polska and Others v Commission EU:T:2017:339, EU:C:2018:756. nuance this ground of (in)action. These cases entailed alleged EU antitrust infringements handled by NCAs, followed by complaints submitted by the investigated undertakings to the Commission, to have the case handled at Brussels, thus aiming to escape (unfavourable) scrutiny in the national jurisdictions. For the Commission to make use of Article 13 of Regulation 1/2003, an NCA must first ‘deal with a case’: i.e. the same alleged infringements, on the same market, and within the same timeframe.Si.mobil (n. 39) [75], [76]; EasyJet (n. 39) [29]; Network Notice (n. 23) [21]. The NCA must review these aspects with a certain intensity, yet without necessarily having to adopt a specific decision on the merits.C.S. Rusu, ‘Workload division after the Si.mobil and easyJet rulings of the General Court’ [2015] 11 Competition Law Review 1. The Commission simply must be satisfied, on the basis of the information available to it when it gives its rejection decision, that an NCA is investigating the case.Si.mobil (n. 39) [50], [75], [77].

A question thus arises regarding the permissiveness of the requirements of Regulation 1/2003, Article 13? First, the Commission may reject a complaint even if NCAs review cases on basis of laws other than competition law: in easyJet, the Dutch NCA applied domestic laws on aviation. The Commission must only be satisfied that NCAs perform a review in light of EU competition law.EasyJet (n. 39) [46]; Trajektna luka Split (n. 39) [27]. In this context, it goes without saying that, while rejecting the case, the Commission cannot rule on the NCA’s arguments, findings, and methodology, this being (review) prerogatives resting with the domestic courts. Second, the Commission may reject complaints even if NCAs rejected the case while using domestic competition law only, without discussing Articles 101 or 102 TFEU, as it was the case in Trajektna luka Split.Trajektna luka Split (n. 39) [28]-[34]. The GC’s argued that Croatian law is the equivalent of EU antitrust law. Consequently, when applying the Automec ‘likelihood of finding an infringement’ test, the Commission, instead of performing its own analysis, could simply rely on the NCA’s reasoning, at least as indication of an infringement of Article 102 TFEU not having taken place. Third, the Commission can also reject a case even if the NCA has rejected the same case due to the expiration of the national law limitation period. In Agria Polska,GC Agria Polska (n. 39) [77], [95]. the GC, confirmed on appeal by the ECJ,ECJ Agria Polska (n. 39) [82]. held that even if the NCA’s rejection does not contain any assessment under EU antitrust law, this does not automatically mean that the Commission must open an investigation. Alternatively, the complainants could claim damages in national courts for any potential EU antitrust infringement. Fourth, the Commission may also reject complaints, if the NCA has rejected the case on priority grounds. In easyJet,EasyJet (n. 39) [26], [34], [39], [40]. the GC argued that the broad scope of Article 13, corroborated with the fact that decisions on priority grounds fit well in the scheme of Article 5(2) (i.e. no grounds for action on NCAs’ behalf), and with the fact that the Commission may dismiss cases even if no NCA deals with a case (i.e. the Automec formula), allow the Commission to remain idle. Otherwise, the Commission would become an appellate body, substituting its role to that of national courts.

Therefore, the threshold for the Commission to reject complaints is quite low. Furthermore, even when faced with claims that NCAs are not well-placed to deal with a case, the Commission is not required to act, neither to check the sufficiency and appropriateness of NCAs’ institutional, financial, and technical means to effectively apply the EU competition rules,Si.mobil (n. 39) [57]; Agria Polska (n. 39) [78]. since the only obligation imposed by Regulation 1/2003 in this respect is ensuring that the effective compliance with the Regulation’s provisions is not frustrated.Council Regulation (EC) 1/2003 of 16 December 2002 on the implementation of the rules on competition laid down in Articles 81 and 82 of the Treaty [2002] OJ L1/1, art 35(1). See also Rusu (n. 18) 38. Lastly, even when the complainants would make a last attempt to trigger to the ‘EU interest’ Automec formula, this would most likely be unsuccessful, because the EU Courts created in the cases discussed above a workaround possibility: when the Commission feels that NCAs are satisfactorily performing the enforcement work, even if no domestic decision on the merits is adopted per se, it can reject complaints even without performing the ‘EU interest’ analysis.Rusu (n. 18) 37.

Concluding, there are lax case allocation requirementsMonti (n. 26) 104 shows, with a reference to Case C-428/14 DHL Express EU:C:2016:27 and to the flour mills cartel (see press release of the Bundeskartellamt from 19 February 2013,

https://www.bundeskartellamt.de/SharedDocs/Meldung/EN/Pressemitteilungen/2013/

19_02_2013_Mühlenkartell.html, accessed 20 May 2019), that the current case allo-cation criteria allow for instances to occur, where various aspects of one and the same practice may be seized by different enforcement agencies. and court-validated instances (essentially connected to fairly light domestic enforcement action) allowing the Commission not to act in the majority of scenarios. This signals a reframed work-sharing formula and a steady Commission deference to the enforcement performed in the domestic jurisdictions.For a critical assessment of the Si.mobil ruling, see P. Figueroa & C. Derenne, ‘Si.mobil v. Commission: Undermining the Effectiveness of EU Competition Law?’, https://eutopialaw.com/2015/10/01/si-mobil-v-european-commission-t-20111-%E2%80%8Eundermining-the-effectiveness-of-eu-competition-law/, accessed 11 December 2018. This undoubtedly sends the message that there is increasing trust in the NCAs’ work, issue which is of particular importance especially when thinking of the efficient use of resources, by avoiding overlapping enforcement activities.Rusu (n. 18) 38. Yet, this setting is only sustainable if a robust conviction exists, that NCAs are truly apt to effectively discharge the EU antitrust enforcement tasks. If, on one hand, that conviction is lacking, the Commission would be able to make use of its power tool to take over a case, based on Article 11(6) of Regulation 1/2003. In none of the cases discussed above such an approach was adopted, or even contemplated. If, on the other hand, the Commission entertains other underlying considerations for relinquishing enforcement action, such as lack of enforcement resources, and if it were not to trust a particular NCA with the investigation of a particular case, it could at least raise the issue within the ECN, to lobby for the reallocation of that particular case. Either way, using an a contrario type of argument, if trust would not be present, enforcement agencies would still retain the means to signal that to their counterparties.

This takes us to the next point of discussion, connected to Directive 2019/1, which aims to boost the NCAs’ enforcement ‘toolkit’ and institutional positioning. We look into whether the Directive solidifies the Commission and NCAs mutual trust relationship or if there is further room to build on this system of flexible allocation of the enforcement tasks in the EU.

4.How is trust reflected in the Directive’s provisions?

4.1.Introductory remarks, objectives and setup of the Directive

The Directive identifies the enforcement areas where further effort is needed to ensure the effectiveness of the antitrust prohibitions, thus developing the findings of the 2014 Communication on Ten Years of Antitrust Enforcement under Regulation 1/2003 and of the 2015/2016 public consultation. The Directive aims to harmonise those domestic rules which ensure that NCAs have sufficient guarantees of independence and resources, and also enforcement and fining powers.Directive, art 1. Such issues also relate greatly to the proper functioning of the Internal Market, as differing domestic enforcement provisions may disadvantage consumers and undertakings alike. To iron out such obstacles for the proper functioning of the Internal Market, the Directive is rightfully based on Articles 103 and 114 TFEU,Directive, Recital 9. For criticism on the use of dual legal basis, see Marco Botta, ‘The draft Directive on the powers of national competition authorities: the glass half empty and half full’ [2017] 38 ECLR 10, 473 et. seq. similarly to the 2014 Private Damages Directive.Directive (EU) 2014/104 on certain rules governing actions for damages under national law for infringements of the competition law provisions of the Member States and of the European Union [2014] OJ L349/1. Directive 2019/1 provides non-exhaustive harmonisation, since only specific institutional, staff and funding matters, and enforcement powers issues are tackled in the text.See also Explanatory Memorandum, 9. Also, in most respects, the Directive puts forward minimum harmonisation techniques. This seems reasonable, given the NCAs’ varying needs, the sensitive nature of some of the proposed items (e.g. sanctioning regimes), and the overall importance of respecting national specificities regarding design, organisation, and funding of the enforcement activities. The Directive thus employs a calibrated approach to harmonisation. For the purpose of our paper, several items dealt with in the Directive have a clear potential to impact on the enforcement trust between the European antitrust enforcers, discussion to which we turn now.

4.2.Institutional design of NCAs

One of the Directive’s aims is to ensure that NCAs are independent when exercising their functions.Directive, art 1(1). Defining ‘independence’ may be difficult, as it can mean different things to different people, and in different contexts.Case C-518/07 Commission v Germany EU:C:2010:125, para 18. See also W. Wils, ‘Competition authorities: Towards more independence and prioritisation?’ [2017] paper presented at New Frontiers of Antitrust 8th International Concurrences Review Conference, 26-27. In general, independence can be said to encompass three facets.Commission Staff working document SWD(2014) 231/2 Enhancing competition enforcement by the Member States’ competition authorities: institutional and procedural issues [2014], http://ec.europa.eu/competition/antitrust/swd_2014_231_en.pdf, accessed 5 March 2019. Firstly, operational independence relates to NCAs’ ability to act independently when enforcing the EU rules (i.e. whether the authority is bound to take instructions from public or private entities). Secondly, organisational independence relates to adequate human resources, meaning the NCAs’ need for sufficient and competent staff members. Thirdly, financial independence is linked to the resources (i.e. budget) NCAs need to perform their tasks.

Independence is important for trust between the different enforcers because NCAs and the Commission collaborate in many instances.E.g. Network Notice (n. 23) [5], [12], [29]. If an NCA cannot rely on its fellow enforcers because their independence is not guaranteed, then these cooperation opportunities become rather meaningless. Yet, this only holds true if there is a positive correlation between independence and effective enforcement. Indeed, it is widely accepted that independence is a prerequisite for effective antitrust enforcement.Commission Staff Working Document SWD(2017)114 Impact Assessment accompanying the ECN+ Directive Proposal [2017], with reference to Case Commission v Germany (n. 59). See also Enrico Alemani et al, ‘New indicators of competition law and policy in 2013 for OECD and non-OECD countries’ [2013] OECD Economics Department Working Papers 1104; OECD, Global Forum on Competition, ‘The objectives of competition law and policy: Note by the secretariat’ [2003] CCNM/GF/COMP(2003)3, 8; UNCTAD, ‘Independence and accountability of competition authorities’ [2008] TD/B/COM.2/CLP/67. Arguments relevant in this regard are that lack of human resources will prevent NCAs from pursuing all possible infringements they come across, and that limited budgets result in less enforcement due to resources unavailability.An empirical study was conducted in 2011, showing that a positive correlation exists between independence and performance: M. Guidi, ‘Does independence affect regulatory performance? The case of national competition authorities in the European Union’ (2011) EUI Working Papers RSCAS 2011/64. The 2015/2016 public consultation revealed that insufficient resources are not merely a theoretically interesting point for discussion, but also a real problem in several Member States.See e.g. replies to the 2015/2016 public consultation ‘Empowering the national competition authorities to be more effective enforcers’ (n. 27), from Bundeskartellamt, 14, the Irish Competition and Consumer Protection Commission, 13-14, the Antimonopoly Office of the Slovak Republic, 16. The reports on the functioning of Regulation 1/2003 have also stressed the importance of guaranteeing the NCAs’ independence.Commission Staff Working Document – Enhancing competition enforcement by the MS’ competition authorities: institutional and procedural issues [2014], http://ec.europa.eu/competition/antitrust/swd_2014_231_en.pdf, accessed 13 November 2018. See also Commission Impact Assessment (n. 62) 27-28. However, in our view, this does not mean that NCAs must not have any external control whatsoever.In literature, different levels of independence were identified, see: Abel M. Mateus, ‘Why should national competition authorities be independent and how should they be accountable?’ [2007] European Competition Journal 3. See also Giorgio Monti, ‘Independence, interdependence and legitimacy: The EU Commission, national competition authorities, and the European Competition Network’ (2014) EUI Working Papers RSCAS 2014/01, 14. In this vein, it is accepted that NCAs should follow general policy guidelines given by a ministry.E.g. Directive, Recital 23. Additionally, a total absence of control could lead to arbitrariness or abuse of powers. The Directive also recognises this, as discussed below. 

Articles 4 and 5 are the Directive’s substantial provisions touching upon NCAs’ independence. Some practical elements stand out when examining these articles. First, Article 4(1) recognises the importance of proportionate account-ability requirements imposed on NCAs.Directive, art 4(1). Recital 22 clarifies that this includes the publication by NCAs of periodic reports on their activities to a governmental or parliamentary body. This is also explicitly mentioned in Article 5(4), which adds a publicity requirement to such reports. For the main part, Article 4 provides safeguards related to the staff and persons who take decisions in NCAs.Directive, art 4(2-4). It provides that these persons shall refrain from taking action which is incompatible with the performance of their duties and exercise of their powers.Directive, art 4(2)(c). However, the exact depth of this obligation is unknown. Some authors have therefore already pointed to Directives and Regulations in e.g. the areas of railway or data protection, to indicate better alternatives to this provision in Article 4.M. Sousa Ferro, ‘Institutional design of national competition authorities: EU Requirements’ [2016] Revista de Derecho de la Competencia y la Distribucion 19; Wils (n. 59). The Directive could have included an obligation for the Member States to lay down by law precise prohibitions on actions and occupations incompatible during and after the staff members’ term of office.E.g. Article 54(1)(f) Regulation 2016/679 on the protection of natural persons with regard to the processing of personal data and on the free movement of such data [2016] OJ L119/1. Instead, there is only a general indication of what could be ‘incompatible with the performance of their duties’, namely when persons taking decisions would be engaged in proceedings which concern businesses by which they have been employed or otherwise professionally engaged, or where those persons or their close relatives would have an interest in those organisations.Directive, Recital 19. Furthermore, Article 4(3) provides rules for the dismissal of persons who take decisions. Unfortunately, the Directive does not indicate a set time period for the mandate that would be long enough to guarantee independence and that would prevent Member States from choosing short renewable terms to diminish the effect of Article 4(3). These Directive provisions provide for a minimum framework all Member States will have to comply with, but can also go beyond. This will most likely increase the trust between NCAs as they can rest assured that the colleagues with whom they cooperate have at least the same minimum safeguards in place for the aforementioned areas of independence.

The parliamentary debate on the Directive provided for a welcome change to the Commission’s proposal. The Directive now includes provisions stating that members of the decision-making body should be appointed through transparent procedures laid down in advance in national law.Directive, art 4(4). Another step in the same direction seems to be the explicit mention in Article 4(5) that NCAs shall have the power to set their priorities and to reject complaints which are not a priority. This is good news for the 15 NCAs that currently do not have this power.Commission Impact Assessment (n. 62) 17. Article 5(1) refers to the Member States’ obligation to ensure at a minimum that NCAs have a sufficient number of qualified staff and sufficient financial, technical, and technological resources, necessary for the effective performance of their duties. However, the Directive does not specify what should be considered ‘sufficient’. This might seem as a drawback, however, in our opinion ‘bright-line harmonisation’ would be inflexible (and undesirable), as NCAs differ greatly in size and exhibit diverse levels of case handling intensity, from one Member State to another. What makes this provision specifically vague is the phrase, ‘in order to effectively perform their duties’. Does this mean that every possible infringement, no matter the impact, should be tackled if the NCA so desires? To our minds, more clarification is needed in this respect. Lastly, the words ‘qualified staff’ in Article 5(1) casts a veil of uncertainty, since the Directive does not indicate how far this provision may reach. Admittedly, it clearly implies that a certain level of expertise is necessary, however, leaves the exact borders of this requirement open to further (case-law?) developments.

An important question to consider is whether the Directive’s provisions surrounding independence create any new obligations for the Member States. If they do not, then it could be said that the Directive will not add much in terms of independence safeguards and therefore to the trust between NCAs. Article 35 of Regulation 1/2003 provides that NCAs’ designation must not jeopardize the Regulation’s provisions. Some authors argue, successfully in our opinion, that Article 35 already implicitly requires a certain institutional design of NCAs.Sousa Ferro (n. 71); Wils (n. 59). Courts could use the effectiveness principle to fill in the blanks that the EU legislator has left, by not providing clearer provisions on the institutional design of NCAs. In VEBIC, this principle was used to argue that Article 35 prohibits national rules which do not allow NCAs to participate, as defendant or respon-dent, in judicial proceedings brought against decisions that the authority itself has taken.Case C-439/08 VEBIC EU:C:2010:739, para 64. The question arises if it was necessary then to lay down provisions on institutional design in the Directive, the answer to which should be affirmative. First, the idea behind the provisions on independence, as becomes clear from the Explanatory Memorandum attached to the ECN+ Directive Proposal, was to give substance to the requirement in Article 35 Regulation 1/2003.Explanatory Memorandum, 4. Second, Article 35 and the principle of effectiveness leave a lot of room for interpretation. Therefore, it was necessary to adopt specific EU rules to define a clearer framework for these requirements. It would be too risky to leave it up to the courts to flesh out what obligations can be derived from Article 35 and the principles of effectiveness, since, until the courts have done so, there is no legal certainty as to how exactly the institutional design of NCAs should look like.Sousa Ferro (n. 71) 28. The stakeholders call for homogenous institutional design in all Member States and this cannot be reached by relying merely on the principle of effectiveness. A Member State could disagree with a judgement that interprets the debatable implications of the principles of effectiveness and equivalence, but is less likely to disagree with the implementation of a Directive in the drafting of which that Member State itself participated.Ibid.

In conclusion, will these new provisions on institutional design of NCAs increase trust between the competition authorities? This is most likely to be the case, since it is crucial, from legal certainty and effective EU antitrust enforcement perspectives, to have institutional safeguards laid down in EU secondary legislation rather than based on EU law principles that are open to interpretation.Ibid. Naturally, it still remains to be seen how the institutional design provisions will be implemented in the Member States. Independence on paper does not necessarily bring actual independence in practice, however, it is still important to regulate, because otherwise the NCAs are left with nothing.Wils (n. 59) 35.

4.3.Investigation powers of NCAs

The Directive also contains provisions dealing with fact-finding/investigative powers of NCAs: inspecting business and other premises (Articles 6, 7), taking statements (Article 8), and summoning representatives of undertakings to appear for interviews (Article 9). This is an area in which the Directive adds value for (at least some of) the NCAs. This is because based on Regulation 1/2003, NCAs were endowed only with decision-making powers (discussed in section 4.4), leaving the investigative tools in the hands of the Member States’ procedural autonomy: the domestic legal regimes provide the fact-finding powers that NCAs could use for tackling alleged anti-competitiveness, of course, while observing the effectiveness and equivalence principles. Yet, not all Member States provided their NCAs with investigative tools sufficient for effective enforcement: for example, the Bulgarian, Italian, and Danish NCAs have no powers to inspect non-business premises, while other NCAs cannot access data stored on servers from other countries, and information on mobile phones, tablets, or laptops.Commission Impact Assessment (n. 62) 16. Such limitations create loopholes which may be speculated by cartelists or dominant undertakings, to the point where, depending on which NCA is competent to act, undertakings may be subject to no or ineffective enforcement, because evidence of anti-competitive practices cannot be collected.Ailsa Sinclair, ‘Proposal for a directive to empower national competition authorities to be more effective enforcers (ECN+)’ [2017] 8 Journal of European Competition Law & Practice 10, 627. This is why the EU antitrust enforcement by NCAs was labelled as rather fragmented.C.S. Rusu & A. Looijestijn-Clearie, ‘Domestic enforcement of EU antitrust and state aid rules – Status quo and foreseen developments’ in A. Looijestijn-Clearie, C.S. Rusu, J.M. Veenbrink (ed.), Boosting the enforcement of EU competition law at the domestic level (Cambridge Scholars Publishing 2017) 7. Furthermore, the inadequacy of investigative powers of some NCAs could have had negative consequences on the functioning of the system of parallel powers for the enforcement of the EU antitrust rules. This is especially relevant for the trust issue discussed in this paper, since NCAs searching for evidence of anti-competitiveness are unlikely to rely on other NCAs (based on Article 22 of Regulation 1/2003) to search for such evidence in their jurisdictions, while knowing that the latter have limited or no powers to do so. In other words, the similarity of investigative powers of NCAs throughout the Member States’ jurisdictions is prone to foster cooperation between NCAs, and consequently the trust in the enforcement work performed by counterparties. It is in this context that the provisions on investigative powers of the Directive must be placed.

Articles 6-9 of the Directive are to a certain extent a codification of the existing EU procedural acquis.Botta (n. 55) 472. Essentially, these articles attempt to match, as much as possible, the Commission’s fact-finding powers in Articles 17-21 of Regulation 1/2003,Rusu (n. 18) 47; Monti (n. 26) 104. and sometimes go further than what this Regulation provides the Commission with. This is so, despite the minimum harmonisation used by the Directive. For example, while Regulation 1/2003 allows the Commission to perform so-called sector inquiries (Article 17), the Directive does not afford such powers to the NCAs. Nevertheless, such powers are conferred by domestic competition laws in some Member States (e.g. the UK, Bulgaria). To our minds, the absence of this tool in the Directive is counterbalanced somehow, by introducing (through the back-door), in Article 8 dealing with requests for information, the NCAs’ possibility of requiring any natural or legal persons to provide relevant information. This is not really a ‘sector inquiry’ power, but nevertheless, it allows NCAs to search for evidence elsewhere than just ‘around’ the undertakings under investigation. Another striking finding relates to Article 9 of the Directive, dealing with the power to summon the appearance for interviews, power which was not present in the Directive’s first draft. However, the EU legislator adopted an interesting, minimum harmonisation construction: NCAs should be empowered to summon representatives of the undertakings under investigation and of other undertakings, to appear for an interview, under the sanction of a fine (discussed in section 4.4). There is nothing in this provision talking about whether you must cooperate with the NCA during the interview, nor about the necessity of consent of the interviewee. Therefore, in our opinion, Article 9 provides no guarantee as to the fruitfulness of interviews as fact-finding tools. Nevertheless, these are matters left in the hands of the Member States, who may choose to regulate further.

Regarding the power to inspect business premises and other premises, Articles 6 and 7 add no surprises. They contain exactly what one would expect, especially when compared to Articles 20 and 21 of Regulation 1/2003, dealing with the Commission’s investigation powers: entering any premises, copying books and records, regardless of the medium on which they are stored, making use of police intervention when needed, the requirement of applying for judicial authorisation, etc. An interesting remark though relates to the level of judicial review the domestic courts may discharge with respect to inspections performed by the Commission (under Regulation 1/2003) and by NCAs (under domestic laws harmonised by the Directive): in the former setting the national courts may review the proportionality, but not the necessity of the inspection, whereas in the latter setting, the domestic court’s review of NCA inspections is not limited to specific aspects in Article 7 of Directive 2019/1. Nevertheless, returning to the Directive’s provisions on the NCAs’ investigation powers, it is precisely these unspectacular provisions which add the most value for those NCAs with limited fact-finding ‘toolboxes’. Speaking of the trust between the NCAs enforcing EU antitrust law, these minimum core investigative provisions create a level-playing field and are prone to further the collaboration and trust between themselves, especially when engaging in extraterritorial evidence gathering. In other words, these provisions give body to Article 22 and the other cooperation provisions in Regulation 1/2003. Concluding, the fact-finding powers in Articles 6-9 are definitely a plus, if not for all, at least for ‘weaker’ NCAs.

4.4.Decision-making and sanctioning powers of NCAs

Articles 10-12 of the Directive deal with the NCAs’ decision-making powers, while Articles 13-16 tackle fines and periodic penalty payments. The former includes the possibility to adopt decisions finding and terminating infringements, interim measures, and commitment decisions. At first glance, Articles 10-12 do not bring many novelties to the table, since Article 5 of Regulation 1/2003 already provided the NCAs with such powers. Nevertheless, the Explanatory Memorandum highlighted the need to ensure that the NCAs’ decision-making powers are fully respected and elaborated on.Explanatory Memorandum, 4. In fact, most, if not all NCAs have already been properly equipped with the power to adopt infringement and commitment decisions, while some NCAs could not adopt structural remedies.Commission Impact Assessment (n. 62) 17, identifies 11 NCAs in this respect. The Directive thus adds value for (at least some of) the domestic enforcers, first because the possibility of adopting such remedies was not present in Article 5 of Regulation 1/2003, and second because Article 10 of the Directive emphasizes the importance of observing the proportionality principle when imposing remedies. Nevertheless, when compared to the Commission’s remedies-related powers under Article 7 of Regulation 1/2003, the Directive does not add much. The same stands for interim measures in Article 11 and commitment decisions in Article 12 of the Directive. A missed opportunity, however, may be identified in the Directive’s text: here we refer to the possibility of adopting negative decisions regarding the EU antitrust rules. Unfortunately, especially when talking about the Commission – NCAs trust relationship, this possibility remains confined to the Commission’s competence (Article 10 of Regulation 1/2003 and Tele2 Polska). The Directive does not build on Article 5 of Regulation 1/2003: Article 10 of the Directive simply states that when NCAs decide that there are no grounds to continue enforcement proceedings and therefore close those proceedings, the Commission should be duly informed about this. Should the Directive have opened up the possibility for NCAs to adopt negative decisions, the NCAs would have been further entrusted to become solid enforcement pillars of the EU rules, and truly ‘European agencies’, for that matter.

Regarding the NCAs’ sanctioning powers, a key aim of the Directive is to ensure that NCAs are able to impose effective, proportionate, and dissuasive fines when Articles 101 or 102 are infringed. This is important, because some NCAs’ investigative and decision-making powers are often without force if they are not backed up by effective sanctions; furthermore, if companies face very low or no fines, depending on which authority acts, this may undermine deterrence and the level-playing enforcement field. Sinclair thoroughly explains how such issues, which essentially create disparities between the sanctioning regimes of the Member States, can affect the level of enforcement of Articles 101 and 102 TFEU.Sinclair (n. 84) 627-628. With a reference to the X ruling of the Court,Case C-429/07 Inspecteur van de Belastingdienst v X BV EU:C:2009:359. she shows that the effectiveness of the penalties imposed by NCAs and the Commission is a condition for the coherent application of the EU competition rules. In this context, the question is then what does the Directive add to Article 5 of Regulation 1/2003, which already empowered NCAs to impose penalties under their domestic laws? As already pointed out, these laws, however, may diverge as to the nature of the fines (administrative, criminal, etc.), or the methodology used for their calculation. Articles 13-16 of the Directive (non-exhaustively) harmonise the NCAs’ powers, thus creating a clear roadmap for imposing sanctions in domestic proceedings, and matching to a certain extent the NCAs’ powers to those of the Commission (Articles 23, 24 of Regulation 1/2003).C.S. Rusu, ‘The Commission’s 2017 EU Antitrust Draft Directive: Addressing the Public Enforcement Fragmentation’, Radboud Economic Law Blog 06/2017, https://www.ru.nl/law/research/radboud-economic-law-conference/radboud-economic-law-blog/2017/commission-2017-eu-antitrust-draft-directive/, accessed 8 January 2019; Monti (n. 26) 104. The Directive creates an interesting route for imposing fines: Article 13 allows Member States to choose between giving the power to impose fines to the administrative NCA (model followed by most Member States) or allow NCAs to apply for the impo-sition of fines in non-criminal judicial proceedings, in front of domestic courts.Sinclair (n. 84) 632. Article 13 also preserves the domestic authorities’ ability to impose sanctions of criminal law nature. The types of infringements for which sanctions may be imposed are once again not surprising: infringements of the EU antitrust substantive rules and a wide range of procedural infringements, which match those already listed in Regulation 1/2003. The odd-one-out is the power to impose sanctions and periodic penalty payments for failures to appear at an interview, referred to in Article 9 of the Directive.Directive, art 13(2)(b), 16(1)(b). This seems to go beyond what the Commission is able to do under Regulation 1/2003, thus giving more bite to the (harmonised and upgraded) investigation powers of NCAs. A similarly odd provision relates to the calculation of the fines that may be imposed. Article 15 of the Directive imposes a so-called ‘minimum maximum’ fine threshold: the maximum amount of the fine that may be imposed by an NCA should not be set at a level lower than 10% of the undertakings’ total worldwide turnover. Certain consequences stem from this approach: first, for those Member States where the maximum amount of the fine to be imposed was previously so low that deterrence was minimal or inexistent, the Directive adds teeth to the antitrust prohibitions; second, for fines imposed domestically, after the Directive’s implementation, the undertakings’ worldwide turnover will be the basis for calculating the level of the fine, whereas the Commission observes the EU-wide turnover under the provisions of Regulation 1/2003; third, the ‘minimum-maximum’ amount of the fine also means that the Directive’s minimum harmonisation would allow Member States to implement a maximum amount of the fine at 15% or 20%, for example. These last two considerations may very well mean that infringers may receive higher fines from NCAs than those imposed by the Commission under Regulation 1/2003. To our minds, this increases the trust placed in the domestic enforcement of EU antitrust law, since the possibility to impose higher fines is prone to increases deterrence. When such deterrence is achieved through fines imposed by NCAs, this inevitably means that domestic action may be thoroughly relied upon, to ensure the effectiveness of the EU antitrust prohibitions. Lastly, along the same lines of increasing trust, one has to appreciate the provisions relating to the principle of parent-subsidiary joint liability (Article 13(5)): for the purpose of imposing fines on parent companies and legal and economic successors of undertakings, the notion of undertaking applies. This means that companies would no longer be able to escape liability by restructuring.

Summing up regarding enforcement trust, the Directive sends mixed signals regarding decision-making and sanctioning powers. On one hand, by providing clear harmonised (albeit minimum) rules on the NCAs’ sanctioning powers, which nevertheless go beyond the Commission’s powers under Regulation 1/2003, the EU legislator shows plenty of trust in what the NCAs may achieve while enforcing the EU antitrust provisions. The ‘minimum-maximum’ fining threshold, calculated on worldwide rather than EU (or national) turnover, allows greater deterrence than before. On the other hand, enforcement trust remains capped in relation to negative decisions, when infringements of Articles 101 or 102 TFEU cannot be established. Therefore, trust can only reach so far, when the coherency of the interpretation of EU antitrust law is at stake.

4.5.Leniency

The Directive also contains rules on leniency (i.e. immunity from and reductions of fines, in exchange for ‘blowing the whistle’ on cartels and cooperating with the investigating authority). Currently, there is no EU-wide harmonised system for leniency programmes. Instead, the Commission applies the soft-law rules laid down in its 2006 Leniency Notice,Commission Notice on immunity from fines and reduction of fines in cartel cases [2006]

OJ C298/17. while the NCAs apply their domestic leniency rules, primarily following the ECN Model Leniency Programme (hereinafter MLP).http://ec.europa.eu/competition/ecn/mlp_revised_2012_en.pdf, as revised in November 2012, accessed 18 December 2018. The MLP encourages Member States to align their leniency systems by providing the minimal elements which should apply to all programmes. It is aimed at soft harmonisation, but does however, not oblige the Member States to have a leniency programme in place. In the DHL judgement, the ECJ ruled that the MLP is not binding on NCAs and that NCAs are free to adopt leniency programmes that are autonomous, not only in respect of other national programmes, but also in respect of the EU leniency programme.Case C-428/14 DHL Express EU:C:2016:27, paras 44, 57.

The MLP provides that an application for leniency to one authority is not considered as an application for leniency to another authority.MLP, 1. It is therefore in the applicant’s interest to apply to all NCAs that may have jurisdiction.Jones & Sufrin (n. 13) 1017. However, given the existing inconclusive case allocation arrangements, potential leniency applicants could be discouraged from applying, due to the discrepancies between existing leniency programmes within the EU (whereas all Member States but Malta have their own leniency systems).Commission Impact Assessment (n. 62) 24 stating that, e.g. summary applications are still not available before some NCAs. In this context, a level playing field, transparency, and legal certainty are key to fostering the incentives to break cartels,Rusu (n. 18) 16. otherwise there is a risk of first, disincentivizing applications, and second, achieving different outcomes for leniency applicants in terms of whether they benefit from immunity from fines or even from fines reductions at all.Directive, Recital 11. Leniency is thus an important tool for detecting and proving cartels,Directive, Recital 50. but predictability is also important for cartelists that consider using the leniency rules; they should be able to predict as accurately as possible what the outcome of their application will be.Leniency Notice 2006, para 6.

Articles 17-24 of the Directive essentially harmonise leniency programmes in the Member States, by codifying the conditions under which NCAs may grant immunity from and reductions of fines,Monti (n. 26) 104. and by approximating the form of leniency statements. However, the Directive also acknowledges that further efforts by the ECN to align leniency programmes might be needed in the future.Directive, Recital 51. It seems that the legislator already acknowledges that compromises were needed during the Directive’s adoption and that therefore some matters were left outside the Directive. An example could be the introduction of a one-stop-shop for markers within the ECN or allowing applicants to submit leniency statements in English or French. Currently, Article 20(3) of the Directive provides that applicants may submit leniency statements in one of the official languages of the Member State of the NCA concerned or in another language of the Union bilaterally agreed between the NCAs. This is rather disappointing, especially considering that the obligation to translate documents may form an obstacle for leniency applicants in terms of time and money. On a different note, the Directive, in essence, brings the MLP soft-law rules into legally binding EU law, which is a step forward in terms of legal certainty for (potential) leniency applicants and for the EU level playing field. Even though some Member States might already have integrated leniency into hard-law, the Directive will, from its entry into force, lay down harmonised and binding rules for all the Member States. In this regard, the only peculiarity would be that the Member States would have leniency rules laid down in binding legislation, while the Commission would still base their leniency policy on soft-law, namely the 2006 Leniency Notice.

Article 21 of the Directive entrusts the Member States to put in place a marker system for immunity applications. This essentially means that an undertaking applying for immunity will be granted a place in a queue for leniency, if it provides further evidence and information within a given timeframe to meet the evidential threshold for immunity. The Member States are not obliged to introduce a marker system for applicants requesting a reduction of fines.Directive, art 21(5). Markers are already widely used in the Member States,OECD, ‘Use of Markers in Leniency Programmes’ (2014), DAF/COMP/WP3(2014)9 14, http://www.oecd.org/officialdocuments/publicdisplaydocumentpdf/?cote=DAF/

COMP/WP3(2014)9&doclanguage=en, accessed 2 January 2019. so Article 21 is not a major novelty. Moving on, Article 22 is particularly interesting, as it provides that NCAs shall accept summary applications from applicants that have applied to the Commission for leniency. An undertaking may make a full leniency application to the Commission, and a summary application to NCAs that may be well-placed to deal with the case. The summary application is a short form and will protect the applicant’s place in the queue if the Commission decides not to take up that case. Even though the non-binding MLP introduced a model for a uniform summary application system, not all Member States followed this possibility.Commission Impact Assessment (n. 62) 24. However, for leniency applicants, summary applications are very important where the alleged secret cartel covers more than three Member States. The Directive ensures that, once the Commission decides not to investigate a case, summary applicants remain free to submit full applications to the relevant NCAs. Their applications will be deemed submitted at the time of the summary application, however only when it covers the same affected product(s), duration of the alleged cartel and territory(ies).Directive, art 22(6). Recital 63 provides that the burden to update the summary applications and to inform the relevant NCAs timely of any changes in scope of a leniency application lies with the applicants.See also https://www.arnoldporter.com/en/perspectives/publications/2018/12/the-new-ecn-directive, accessed 2 January 2019.

All in all, these new rules will make leniency a much more reliable and predictable process, which will increase the desired legal certainty of leniency programmes. In terms of trust, the NCAs and the Commission are prone to have more confidence in each other now that harmonised leniency rules will apply in all Member States.   

4.6.Mutual assistance

Articles 24-29 of the Directive provide rules on mutual assis-tance between the NCAs. These provisions are particularly helpful to further the actions under Article 22 of Regulation 1/2003: NCAs may perform inspections or interviews under their national laws on behalf of other NCAs. Regulation 1/2003 introduced a system of parallel powers for the enforcement of the EU antitrust provisions based on close cooperation within the ECN. However, the Explanatory Memorandum underlines that this cooperation becomes undermined if there are still NCAs that do not have adequate fact-finding tools.Explanatory Memorandum, 3. Because differences in national procedures exists, it may become less attractive to cooperate and trust another NCA to carry out inspections on its territory and under its own national law, if it has few or no investigation powers. By harmonising the national fact-finding rules, Article 22(1) of Regulation 1/2003 will be given more substance and mutual trust between NCAs will increase. From that perspective, the provisions on mutual assistance in the Directive are a logical and necessary addition to bring the cooperation between NCAs even further, thus boosting trust between these authorities.

Chapter VII of the Directive refers to the terms ‘applicant authority’ and ‘requested authority’, meaning a national competition authority.Directive, art 2(20,21). Therefore, all the rules regarding mutual assistance refer to NCAs among each other, and not to NCAs and the Commission. It is peculiar that Recital 68 acknowledges that close cooperation is required among NCAs and between NCAs and the Commission, while all the provisions on mutual assistance only apply among NCAs. Either way, once the Directive is implemented, NCAs will be able to request other NCAs for the enforcement of decisions imposing fines and periodic penalty payments. This possibility in essence comes down to NCAs recognising each other’s decisions,Directive, art 26(3) provides that the request may only be made for a final decision. an issue which was specifically mentioned as highly desirable during the 2015/2016 public consultation.See e.g. reply to the 2015/2016 public consultation ‘Empowering the national competition authorities to be more effective enforcers’ (n. 27), from the Romanian Competition Council, 12. While NCAs cooperated and entrusted each other with investigative work based on the provisions of Regulation 1/2003 and through the means provided by the ECN, when it came to enforcing final antitrust fining decisions issued by NCAs from other Member States they were faced with administrative difficulties arising from the domestic provisions. Embedding the possibility to recognise and enforce other NCAs’ decisions directly in the Directive remedies this drawback. To our minds, this is a major novelty which solidifies the existing trust between NCAs, especially when compared to the informal system of collaboration which characterises the ECN.Botta (n. 55). What this means, in our opinion, is that a certain level of harmonisation is created by Article 26 Directive. The article does not bring about ‘standard setting’ for all Member States, but instead applies a lower threshold. It requires a specific degree of integration: the NCAs apply their own national rules, but at the same time, in principle, must recognise fining decisions of other NCAs, meaning that a decision taken in one Member State will also apply in other Member States.

Article 27 contains a number of general principles of cooperation and is a novelty that has been added to the initial Commission Proposal. One of these general principles is that requests under Articles 25 or 26 of the Directive may be rejected by the requested Member State under certain circumstances (e.g. manifest non-compliance with public policy requirements). Furthermore, Article 27(2) refers to a ‘uniform instrument’ that is to be used when a Member State requests another Member State under Articles 25 or 26 Directive. This uniform instrument shall constitute the sole basis for the enforcement measures to be taken by the requested NCA.Directive, art 27(4). Its use will most likely create plenty potential to increase trust among NCAs. Article 33(2) of the Directive, which falls under the Chapter on general provisions, provides room to even further this trust in the future. It provides an indication and an opportunity for the ECN to establish best practices and recommendations on several matters, of which mutual assis-tance is one. Concluding, the provisions on mutual assistance will provide a positive boost for increasing enforcement trust among the antitrust enforcers in the EU.

5.Conclusion

Summing up, the Directive shows a high degree of trust in the NCAs. Testimony to this is the fact that some of its provisions empower NCAs beyond the level of empowerment the Commission has under Regulation 1/2003. This trust between the ECN members may be further strengthened. The key however relates to the Directive’s implementation in the Member States and to the new rules’ actual practical application.

Regulation 1/2003 provided the initial basis for a trust relationship to develop, by sharing the power to enforce the EU antitrust rules between the Commission and the NCAs. However, that trust was not unlimited, and was ‘lightly controlled’ under the effectiveness and equivalence EU law principles. The Directive provides much needed handles in this regard. It moves away from the effectiveness and equivalence safety-net, by harmonising the domestic rules on institutional and procedural matters.

One remaining question is whether the new rules will benefit all NCAs. Most likely not, as only those Member States that currently have institutional or procedural shortcomings will truly directly benefit from the Directive. Furthermore, there is even room to move beyond the Directive, as there are still some loose ends. For example, the Directive misses the opportunity to bring case allocation from soft-law into a legally binding framework; it could have also ensured that NCAs can adopt negative decisions, as true ‘EU agencies’. Lastly, Recital 51 of the Directive states that effort should be put in further aligning leniency programmes. This signals that EU antitrust enforcement is an ongoing, developing process. The Directive is nevertheless a valuable piece of legislation. It can be a catalyst to further NCAs’ powers and thereby the EU antitrust enforcement trust between them and the Commission.