Private Enforcement of the EU Rules on Competition – Nullity Neglected

Private enforcement of the European Union’s rules on competition (Arts. 101, 102 TFEU) has become prominent as a counterpart to their public enforcement. Mostly, it is identified with tort actions brought under EU-harmonized national law by individuals claiming compensation for the harm suffered from anticompetitive agreements or practices. However, claims for compensation represent imperfect sanctions for the infringement of the competition rules because they are brought only once the damage is done and at a time when the conditions of competition may have changed. Typically also, such private actions are no equivalent or complement to administrative enforcement, but are largely dependent on it (follow-on actions). In addition, bringing them is attractive only if the damage suffered is considerable, sufficient evidence available, and the defendant solvent enough. Therefore, this paper revisits the first line of private enforcement, which is enforcing the nullity of anticompetitive agreements as provided for directly by primary Union law in Art. 101(2) TFEU. Nullity was a much-discussed issue under the authorization regime of Reg. 17/62, the first regulation implementing the enforcement of the competition rules, but has become somewhat neglected as a sanction since Reg. 1/2003 changed the enforcement system. Yet, it is precisely under the regime of immediate and direct applicability of both Arts. 101(1) and 101(3) TFEU, which Reg. 1/2003 reestablished, that the potential of nullity as a sanction of anticompetitive agreements could be fully activated. Such active use of invalidity challenges may lead to redefining the interface between EU law and national contract law, which is the line of severability of the innocent parts of a restrictive agreement from its anti-competitive parts. It should also result in reassessing the legal fate of follow-on transactions concluded by a party to an anticompetitive agreement with third parties, and it should bring abusive contracts within the realm of the nullity sanction that dominant firms impose on third parties. The guiding principle for such general reappraisal of the nullity sanction must be to bring its purpose fully to bear, which is to facilitate exit from anticompetitive agreements or from (abusive) contract clauses with a view to reopening competition and/or to allow the renegotiating of a transaction in terms of undistorted competition. This may mean that only the party whose freedom of competition is restricted may claim nullity.

transactions concluded by a party to an anticompetitive agreement with third parties, and it should bring abusive contracts within the realm of the nullity sanction that dominant firms impose on third parties. The guiding principle for such general reappraisal of the nullity sanction must be to bring its purpose fully to bear, which is to facilitate exit from anticompetitive agreements or from (abusive) contract clauses with a view to reopening competition and/or to allow the renegotiating of a transaction in terms of undistorted competition. This may mean that only the party whose freedom of competition is restricted may claim nullity.
Keywords EU competition rules Á Private enforcement Á Nullity under Arts. 101(2) and 102 TFEU Á Severability of non-restrictive clauses Á Follow-on transactions Á Actions for damages

Introduction
In the early seventies, when Jochen Pagenberg and I met as young members of the research staff of the Max Planck Institute for Intellectual Property in Munich 1 and began working for the IIC, private enforcement 2 of the EU rules prohibiting agreements and concerted practices in restraint of competition (Art. 85 EEC Treaty 3 ) and the abuse of their power by market dominating enterprises (Art. 86 EEC Treaty) was only at its very early beginnings. Hardly anybody expected it ever to leave the long shadow of administrative enforcement. Based on Art. 87 EEC (now Art. 103 TFEU), the ''First Regulation Implementing Articles 85 and 86 of the Treaty'', 4 while recognizing the power of Member States' authorities to apply the prohibition rules of Art. 85(1) and Art. 86 EEC as a matter of their direct effect, entrusted the Commission with broad powers to enjoin by administrative decision those anticompetitive practices that violate Art. 85 or 86 EEC. However, it withheld from national competition authorities the power to apply Art. 85(3) EEC (now Art. 101(3) TFEU), 5 the broad exemption rule that forms the counterpart of the general prohibition of restrictive agreements, by reserving such power exclusively for the 1 Then the Max Planck Institute for Foreign and International Patent, Copyright and Competition Law. 2 The common juxtaposition of ''private'' and ''public'' enforcement corresponds to U.S. terminology. It does not fit well into continental civil law systems where public enforcement corresponds to administrative enforcement and private enforcement rests on remedies of civil (or ''private'') law. As such, it is not ''private'' but as much in the public interest as is administrative enforcement. 3 The EEC (European Economic Community as established by the Treaty of Rome of 25 March 1957) became the European Community (EC) in 1992/93 as part of a ''European Union'' (EU), which the Treaty of Maastricht created. In 2009, the Conference of Lisbon resulted in consolidated versions of the Treaty on European Union (TEU) and the Treaty on the Functioning of the European Union (TFEU); see OJEU 2016 C 202, 1, 47. While the wording of the competition rules was never affected by these changes, they were renumbered from Arts. 85,86 EEC to Arts. 81,82 EC Treaty and Arts. 101,102 TFEU. 4 Regulation No.17/62 of the Council of 6 February 1962, OJEC 1962 Commission. Since no major agreement 6 could benefit from the exemption unless it first had been notified with the Commission and, thereupon, formally authorized by it, administrative control was systematically broad and amounted to what was called a ''principle of prohibition subject to a reservation of administrative authorization. '' 7 In practical terms, this meant that the basic sanction for anticompetitive agreements, their ex lege or automatic nullity (Art. 85(2) EEC, now Art. 101(2) TFEU), would remain in legal suspense until the Commission had taken a decision. Thus, nullity would operate effectively only in regard of non-notified agreements or of agreements for which an exemption had been finally denied, and, possibly, in respect of agreements for which an exemption could not seriously be expected. 8 Thus, private enforcement was limited from the outset. It was so also because tort liability entitling one to compensation for damages suffered as a result of an anticompetitive practice, be it uni-, bi-or multi-lateral, was not provided for in the Treaty or in Reg. 17/62. Liability was a matter of Member States' domestic tort law. Consequently, it differed nationally, and, for instance in Germany, 9 was controversial as to its availability and personal reach. In addition, at least as a practical matter, any liability could be enforced only once the incompatibility of a restrictive practice with Art. 85(3) EEC (now Art. 101(3) TFEU) had been confirmed administratively or was otherwise evident enough to risk investment in litigation. 6 Arts. 4 and 5 Reg. 17/62, which established the notification requirement, distinguished between mandatory notification of major agreements and voluntary notification of minor agreements (such as agreements between parties of one Member State only or bilateral agreements establishing simple distribution systems or simple license agreements; in addition, some ''innocent'' forms of horizontal cooperation, such as collective standardization, joint R and D, and specialization agreements, did not need to be notified). 7 Since at the time the German Act Against Restrictions of Competition (ARC) of 1957 provided for the most developed ''antitrust law'' of the then six Member States of the EEC, Reg. 17/62 followed its model to some extent, a general prohibition of both horizontal and vertical agreements (Secs. 1 and 15, respectively) accompanied by a notification-based system of limited exceptions for horizontal agreements (Secs. 2-8) and broad exceptions for vertical agreements , with notification required for vertical price-fixing while other restraints were subject to ex post control only. License agreements benefitted from a special regime (Secs. 20,21). Market-dominating enterprises were subject only to an ex post administrative control for abuses (Sec. 24). However, exclusionary practices and discrimination were subject to a specific rule of prohibition, Sec. 26(2). The Act has been amended repeatedly; in its current version (of 26 June 2013, BGBl. I., 1750, as last amended on 19 January 2021, BGBl. I 2021,2) it almost entirely follows the EU-model as regards both its substance and the enforcement regime of Reg. 1/2003 (see infra text following n.10). A remaining particularity of importance is the extension of control to exclusionary or discriminatory practices by firms enjoying relational or simply superior market power, Sec. 20 ARC; see infra note 91. 8 For the status of agreements concluded after the entry into force of Reg. 17/62 and the early doctrinal controversies, see Oberdorfer et al. (1971), § § 73 et seq., 301 et seq.; Mailänder (1972), Art. 85, annot. 80 et seq.; Waelbroeck (1972), p. 156 et passim; extensively Mestmäcker (1974), p. 249 et passim, 258 et passim;Ullrich (1971), p. 144 et passim. For the legal uncertainty resulting from CJEU, 6 February 1973, case 48/72, Brasserie de Haecht / Wilkin & Janssen (de Haecht II), Rep. 1973, 77, 86 et seq., ECLI:EU:C:1973:11, at paras. 10 et seq., see infra n.11 and text at 2.2.1.1. 9 Under Sec. 35(2) ARC 1957 a claim to compensation would be given only to those victims of an anticompetitive practice that the infringed rule of prohibition is intended to protect individually. This criterion could work differently with regard to both the different rules of prohibition and the market position the claimant has in relation to the anti-competitive practice; see Mestmäcker (1974), p. 574 et passim; with respect to infringements of the ARC see Emmerich (1981) § 35 annot. 10 et passim; for other national laws see Waelbroeck (1972), p. 177 et seq.
After 40 years of operation, the rules of the game changed fundamentally when, in 2003, the then European Community replaced Reg. 17/62 with Regulation (EC) No. 1/2003 on the implementation of the rules on competition. 10 By this reform regulation, the European Community entirely abandoned the system of prior notification/authorization of restrictive agreements qualifying for an exemption and, most importantly, withdrew the Commission's privilege of having exclusive competence to apply Art. 81(3) EC Treaty (now Art. 101(3) TFEU). It thus made the exemption work again as an exception by law. 11 This meant not only that the parties to the agreement had to themselves assess the validity of their agreement and that the Commission, being relieved of the notification/authorization workload, could re-focus its administrative activity on actual enforcement. 12 Also, Reg. 1/2003 opened a direct path to private enforcement in that the validity/nullity of restrictive agreements was no longer in any suspense. Depending on whether or not they qualified under Art. 81(3) EC Treaty (now Art. 101(3) TFEU) agreements were either valid or invalid, and, in the latter case, vulnerable to attack. Parties to the agreement could more easily disregard or withdraw from it; third parties did not have to respect it. Instead, they could directly and immediately hold the parties to the agreement liable under applicable national tort law for damages suffered from the anticompetitive effects of the agreement. Thus, private enforcement became a more generally available alternative to administrative enforcement, particularly as its potential was considerably enhanced by Directive 2014/104/EU on certain rules governing actions for damages under national law for infringements of the competition law provisions of Member States and of the European Union (so-called Damages Directive). 13 Based on Arts. 103 and 114 TFEU, this Directive does not merely harmonize Member States' laws as they apply to ''antitrust torts'' but considerably strengthens the enforcement of tort claims by obliging Member States to ensure the full recovery of damages suffered and profits lost and to provide for procedural rules, in particular on access to evidence, that facilitate effective enforcement. 14 10 OJEC 2003 L 1, 1. For the reasons for the reform see Commission (1999,2000). 11 See Art. 1(1), (2) and Art. 6 Reg. 1/2003, Art. 6 confirming that national courts have the power to (fully) apply Arts. 81 and 82 of the Treaty (now Arts. 101, 102 TFEU). The provisions are but a late consequence of the de Haecht II judgment of the CJEU (supra note 8) that confirmed both the retroactive effect of Commission decisions refusing to apply Art. 85(3) EEC Treaty (now Art. 101(3) TFEU) to an agreement and national civil courts' authority to apply both Art. 85(1) and (3) EEC Treaty. 12 For the Commission, its power to terminate arrangements and practices that infringe Arts. 101, 102 TFEU (i.e. that come within Art. 101(1) and do not meet the conditions of Art. 101(3) TFEU or that constitute an abuse of market power, Art. 102 TFEU) became the central piece of its authority. Related to this power is the enforcement option offered by Art. 9 Reg. 1/2003 (termination of infringement procedures by accepting from the alleged infringer binding commitments to stop or undo the alleged anticompetitive practice or effects). In case it is in the public interest, the Commission may also issue a negative clearance (non-applicability of Art. 101(1) or Art. 102 TFEU) or a decision confirming that an agreement qualifies for the Art. 101(3)TFEU exception, Art.10 Reg.1/2003. Since by its Arts. 3, 5 and 11 et seq. Reg. 1/2003 obliges national competition authorities to always also apply the EU competition rules when they enforce national competition law, the Commission's workload is further reduced so as to allow it to focus on major and/or on non-routine cases. 13 Directive of the European Parliament and of the Council of 26 November 2014, OJEU 2014 L 340, 1. 14 For details see infra 2.3.1.
Jochen Pagenberg preferred the auspices of the liberal profession of an attorney to a university career, but always remained associated with academia, not least by directing the IIC over so many years. Therefore, dedicating to him some academic considerations on the enforcement of the law may be justified, the more so as private enforcement is where the competition rules and the protection of intellectual property meet frequently enough, 15 albeit sometimes in conflict. By coincidence, only a few months ago, the Damages Directive, which aims at strengthening private enforcement, underwent its first evaluation by the European Commission. 16 However, rather than undertaking an assessment of the workability of the Directive in its practical detail, this paper's concern will be with the more general question whether private enforcement of the EU's rules on competition is in itself systematically well balanced. To a considerable extent, private enforcement of EU competition law rests on Member State law. This asks for taking an approach of comparative law. Unfortunately, the perspective of this paper must be limited to that of German law, in particular to the substantive and procedural rules of the Act  Wollenschläger et al. (2020), Chapters 4-8. 18 Art. 106 TFEU is two-faced. In its para. 1, it obliges Member States that grant special or exclusive rights to their (public or private) undertakings to abstain from any measures inducing these undertakings to violate the rules of the Treaty. In its para. 2, it provides for an exception to Arts. 101 and 102 TFEU insofar as the application of these prohibitions would obstruct the performance of the services of general economic interest that have been entrusted to an undertaking. Since, as such, public undertakings are subject to Arts. 101, 102 TFEU, the general rules on private enforcement apply, and they apply where the criteria of the Art. 106 (2) TFEU-exception are not met. 19 The legal strength of the rule becomes more apparent in the French version: ''Les accords ou décisions interdits en vertu du présent article sont nuls de plein droit''. no equivalent rule of ''contract'' law in the Treaty regarding the ''private'' or, for that matter, civil law consequences of an abuse of market power, be it by ''agreement'', nor is there any primary law provision on the compensation of damages suffered by third parties, victims of an infringement of Arts. 101 or 102 TFEU. However, it follows from the direct effect Arts. 101 and 102 TFEU produce within the legal order of Member States that the prohibitions laid down by these provisions form part of the framework rules governing the national systems of contract and tort law. Thus, their infringement may be sanctioned pursuant to the rules that generally apply to this kind of prohibitions under contract and tort law. 20 Accordingly, a two-level system of private enforcement has been developed wherein by interpreting Union law the Court of Justice of the EU (CJEU) has set and may set the principles governing private enforcement while national law, as applied by Member States' courts and elaborated on by doctrine, caters to the implementation of these principles. The degree of interdependence and interaction between these two levels varies according to the extent Union law predetermines the terms of private enforcement law. Therefore, the following overview begins by presenting the sanction with which the Treaty itself threatens agreements infringing its Art. 101, namely their nullity. (1) TFEU became exposed to the risk of being held invalid ab initio by civil courts of Member States. 24 Although handed down only in reaction to the specific problems created by the notification mechanism of Reg. 17/62 these rulings of the CJEU remain significant for today's enforcement system because of the importance that the Court attaches to the prohibition laid down in Art. 101(1) TFEU and the ensuing need to sanction it with severity. 25 The weight given to Art. 101 TFEU and its strict enforcement is evidenced by the fact that, despite the broad powers Art. 103 TFEU confers upon the EU legislature with regard to the implementation of Arts. 101 and 102 TFEU, the Court did not allow Reg. 17/62 to curtail the direct applicability of Art. 101 TFEU or of its nullity rule. 26 It results from both the importance the Court generally attributes to the direct effect of the Treaty rules, in particular to those governing Member States' integration into an Internal EU Market, and from the particular role the rules on competition are supposed to play in that regard. 27  Reg. 17/62), the agreements really were ''in suspense'' and could fall on either the side of Art. 101(1) or on that of Art. 101(3) TFEU. Therefore, both in de Haecht II (supra note 22) and in BRT/SABAM (supra note 23) the Court ''invited'' national courts to either submit questions of interpretation of Art. 101(3) TFEU to the CJEU pursuant to Art. 177 EEC (now Art. 267 TFEU) or, if the Commission already had initiated its proceedings, to suspend adjudication, except if the agreement clearly did not come within the ambit of Art. 101(1) TFEU in the first place or obviously did not qualify for Art. 101(3) TFEU. However, this was an unsatisfactory approach to the ''contradictions, lacunae and obscurities'' (AG. Roemer, Opinion of 13 December 1972 in case 48/72de Haecht II -Rep. 1973, 90, 98 (ECLI:EU:C:1972:122) of the system of control. The Art. 267 TFEU-route is no option, because the problem of Art. 101(3) TFEU is its application, not its interpretation. The alternative option, suspension of judicial proceedings until the Commission has terminated its administrative procedure, meant waiting for years for an uncertain outcome. Therefore, a sort of a litigation truce came to be observed. It created so much de facto certainty that industry would later on oppose the direct applicability of both Art. 101(1) and Art. 101(3) TFEU as provided for (by way of clarification) by Reg. 1/2003. 25 Thus, in its de Haecht II judgment (supra note 22 at para. 5) the Court stated that it is the intention of Art. 85(2) of the Treaty (now Art. 101(2) TFEU) ''to attach severe sanctions to a serious prohibition'' (in the authentic French version: ''destiné à sanctionner avec sévérité une interdiction importante''). In fact, in de Haecht II, the Court departed from the opinion of its Advocate General Roemer (supra note 24) and from its ruling in Bilger/Jehle (supra note 21), which both were in favor of according the agreements some (provisional) validity. The Court reaffirmed the importance it attaches to Art. 101(1) and (2)  provision on free trade that, as early as 1963, motivated the Court 28 to establish the direct applicability of Treaty rules as a foundational principle of the EU. This principle it then applied successively to all four freedoms of free movement of goods, services, capital and establishment with a view to activating the interest and willingness of individual market actors in directly enforcing them by way of an attack on or a defense against state-induced barriers to integration. 29 It is also in this perspective that the rules on competition and their strict enforcement reveal all of their importance. They ensure both market integration through free competition across state borders 30 and the well-functioning of the system of undistorted competition upon which the legitimacy and the acceptance of the Internal Market and much of the EU's policies rest. 31 2.2.1.2 The rationale: A sanction and an invitation to exit As Art. 101(2) TFEU is meant to sanction violations of the prohibition of anti-competitive agreements severely, 32 the meaning and reach of the nullity of such agreement need to be understood in terms of competition law and applied so as to effectively satisfy its policy objectives. This approach raises no problems as regards determining the meaning of nullity as such. It is generally accepted that an agreement that, due to Art. 101(2) TFEU, has to be considered void may not give rise to any claims to performance or to compensation for non-performance or to a defense based on nonperformance. 33 It is also beyond doubt that Art. 101(2) TFEU may not be contracted away or around by providing for penalties or for obligations to give notice of 28  (2) renders … agreements automatically void. Such nullity is therefore capable of having a bearing on all the effects, either past or future, of the agreement or decision''. For details see Schmidt (2019), Art. 101(2) AEUV, annot. 13; ibid., Appendix 2 VO 1/2003, annot. 10. In Be´guelin, the Court has also held (at para. 29) that an agreement that is void under Art. 101(2) TFEU ''cannot be set up against third parties'' (''n'est pas opposable aux tiers''); to the same effect CJEU, 20 September 2001, case C-453/99, Courage/Crehan, Rep. 2001I 6297, 6322, ECLI:EU:C:2001. This complementary part of the general rule of nullity is important mainly with regard to anticompetitive non-compete clauses and distribution agreements for which a third-party effect may be claimed under national law (see e.g. Chagny (2004), p. 436 et passim). Such third-party effects have attracted much attention in the early decades of market integration, but will not be followed up here again as the principles are settled by now. termination or of withdrawal from the agreement. The reason is not so much the doctrinal qualification of Art. 101(2) TFEU as a rule of public policy 34 as the fact that any such stipulation would frustrate the purpose of Art. 101(2) TFEU, which is to preserve entirely the parties' ex ante freedom and ex ante competitive position by negating ab initio any binding force of the agreement. In Courage v. Crehan, 35 the CJEU relied on precisely this absolute effect of Art. 101(2) TFEU for ruling that a party to an anticompetitive agreement may claim compensation by the other party for the harm suffered from the agreement. Likewise, a party to an anticompetitive agreement, and be it the party initiating it, may not by an ''unclean hands'' counterdefense be precluded from raising antitrust claims or defenses against the other parties. 36 Also, a party to the anticompetitive agreement is not precluded from claiming restitution of performance made under the agreement. 37 The reason supporting this seemingly far-reaching effect of Art. 101(2) TFEU is that, as a sanction, nullity is not simply the result of the anticompetitive agreement being outlawed by Art. 101(1) TFEU. Rather, there is a competition policy rationale underlying it, which is to provide a shield against and an escape from the pressure exercised inside a cartel or by virtue of such market power as is necessary to induce a party to enter into restrictive covenants of a business transaction. The aim of Art. 101(2) TFEU, therefore, is to allow at any time an unconditional exit out of the restraint and to facilitate re-entry on the market or renegotiation of a transaction on terms of free and undistorted competition. Thus, it is generally recognized that the 34 See the EcoSwiss ruling of the CJEU, supra note 25, and note that, as a result, the national judge has to apply Art. 101(2) TFEU of its own motion (ex officio  37 In its Courage/Crehan-ruling (supra note 35, at para. 30) the CJEU, while admitting damage claims between parties to an unlawful agreement, has left the exceptional application of an unclean-hands defense to determination by national law. The problem with this Delphic ruling is that under national law the matter is controversial and apparently not dealt with consistently by courts, see Schmidt (2019) nullity sanction purports to facilitate and even to invite exit from the cartel. 38 Nullity is, indeed, well suited as a remedy to horizontal agreements, in particular to hardcore cartels of the price-or quantity-fixing type and the like. Any party to such agreement may claim or rely on it in its own interest and, more importantly, as a matter of halting or at least limiting the general harm the agreement does to the economy. In the case of vertical restraints that characterize or accompany distribution systems or license agreements, the anticompetitive effects typically are more ambivalent. Therefore, the reach of the nullity sanction needs to be flexibly adapted to the twofold goals that the competition rules pursue in this context, namely, on the one hand, that of keeping markets and consumer choice open, and, on the other, that of protecting the individual freedom of the party upon which a restriction is ''imposed'' or which is otherwise caught in a system of vertical restraints. 39 Depending on the circumstances, this may mean that each of the parties may invoke the invalidity of their ''arrangement'' or only the party whose freedom of competition is restricted.

The reach: Severability and follow-on transactions?
It is in the light of the competition policy rationale of Art. 101(2) TFEU that the consequences of the nullity of an anticompetitive agreement need to be determined. 40 This is all the more important as there is a divide between the nullity verdict following from EU-wide uniform Union law, on the one hand, and, on the other, its consequences, which are a matter for determination by the domestic laws of Member States. Since, by definition, any agreement infringing Art. 101 TFEU affects interstate trade in the EU, its nullity needs to operate unitarily across state boundaries so as to equally cover and, hopefully, undo or limit the anticompetitive arrangement and its effects everywhere. However, as it is Member States' laws that are called upon for determining the consequences that the nullity sanction may entail as regards the ''innocent'' parts of a complex agreement or as regards so-called implementing and/ or follow-on transactions, diversity of legal principles and views will be the rule. 38 For the ''natural'' instability of cartels and the economic conditions determining it, see Kerber and Schwalbe (2020) . Given the broadly varying conditions of competition and restrictive agreements, a case-by-case approach is recommendable. A recurring criterion will be whose anticompetitive interests are served by a restriction, and whether these are mutual or unilateral (in which case the ''victim'', whose freedom is restricted, may need protection). At any rate it is not the equities between the parties that matter or their unclean hands, but how the goals of competition law will be met best. 40 In its Courage/Crehan ruling (supra note 35, paras. 30 et seq.), the CJEU has limited itself to stating with regard to damage claims, which are a matter of national law, that Union law does not stand in the way of national law providing for a flexible treatment of damage claims raised by a party to the anticompetitive agreement against another party. However, the Court seems to be in favor of such competition-oriented flexibility. 41  Therefore, it is important that the Court's repeatedly confirmed ruling 41 be purposively implemented according to which it is Union law that determines the reach of the nullity sanction with a view to fully covering the agreement's anticompetitive operation and effects everywhere, whereas national law may come in only for those of its parts that are entirely outside Art. 101 TFEU. It is for ensuring both the uniformity of application and enforcement of the nullity sanction and the full respect of the competition policy rationale of Art. 101(2) TFEU that the question of the severability of innocent parts of an agreement from its anticompetitive substance must be a matter also of Union law while only the fate of the truly innocent parts may be determined by national law. More precisely, severability marks an interface between Union law and national law, and, thus, may and must always be looked at from both sides. 42 For example, where anticompetitive agreements are not concluded as such, i.e. on a stand-alone basis, but as part of or within the context of a broader business transaction, they may serve as the support of the transaction or they may be supported by the transaction; in either case, the entire transaction is a matter of concern for Union law of competition. More generally, severability will not exist for any parts of the agreement that, although not themselves an expression of a restriction of competition, are functionally linked to or supportive of the anticompetitive object, operation or effect of the agreement. 43 Whether the remainder of the agreement is self-contained enough to survive will depend on its nature as assessed under the possibly diverging rules and judicial application of national contract laws, some being particularly inventive as regards judicial intervention in favor of maintaining the agreement alive in reduced form. 44 Another source of diversity is the distinction between those cartel-related transactions that implement the cartel (or other forms of restrictive arrangements, 42 Accordingly, questions regarding the interpretation of the severability threshold may be -and eventually need to be -submitted to the CJEU as a matter of ensuring uniformity of the law (Art. 267 TFEU). Most authors, while recognizing the interface, tend to look at it from the side of national law; see Schmidt (2019), Art. 101(2) AEUV, annot. 22, 23; Jaeger (2020), Art. 101(2) AEUV, annot. 20. 43 See Jaeger (2020), Art. 101(2) AEUV, annot. 20 et seq. The theoretical distinction between the issue of severability of innocent parts of an agreement from its anticompetitive parts on the one hand, and, on the other, the issue of whether the remaining innocent part may survive or may be made to survive will not always work in practice. For instance, a non-attack clause in a patent license agreement (example by Schmidt (2019), Art. 101(2) AEUV, annot. 22) will be severable on the face of it, and the remainder may be retained as a regular license transaction. However, the more challengeable the licensed patents are, the less likely it will be that the patentee would have granted the license in the absence of the particularly restrictive no-challenge clause. Again, as mentioned in the text, the problem is not only to determine which contractual clauses support the restrictive arrangement, but also, conversely, to determine whether the anticompetitive parts constituted the support of the transaction. 44 See for German law Schmidt (2019) so-called ''implementing agreements'') and those that the parties to an anticompetitive agreement enter into with third parties on up-or downstream markets in observance of their collusion or concertation 45 (so-called ''follow-on transactions''). 46 The former are undoubtedly caught by Art. 101(2) TFEU, the latter possibly not, but the delimitation line between the two is not always clear 47 and the validity of follow-on transactions a matter of persistent controversy. 48 2.2.2 Art. 102 TFEU 2.2.2.1 Asymmetric invalidity Abuses of market power as prohibited by Art. 102 TFEU may be strictly unilateral, such as absolute or discriminatory refusals to deal that increasingly bring claims to access to the fore of private enforcement, typically as a matter of relief from tortious conduct. 49 However, frequently enough, abusive conduct by market-dominating firms will take the form of restrictive contracts, particularly of the kind listed by Art. 102 lit. a)-d) TFEU. Since Art. 102 TFEU is not complemented by a rule of nullity, as is Art. 101 TFEU, any sanctions for its infringement are a matter of Member States' laws. Such national laws must give full effect to the directly applicable prohibition that Art. 102 TFEU provides for. It is 45 Contrary to a widely held view (e.g. Schmidt (2019), Art. 101(2) AEUV annot. 3; Jaeger (2020), Art. 101(2) AEUV, annot. 18), transactions based on a concerted practice may well be held void not on the basis of Art. 101(2) TFEU, but in consequence of the prohibition rule of Art. 101(1) TFEU (see van Damme (1980), p. 209. After all, follow-on transactions are concluded once the anticompetitive agreement is put into operation, and since such agreement is void by virtue of Art. 101(2) TFEU, its putting into operation is tantamount to a concerted practice (albeit not in a technical sense). Therefore, there is no reason why follow-on transactions based on concerted practices properly speaking should not be treated the same as any follow-on transaction, the less so as it is by the follow-on transactions that the coordination becomes a concerted practice; comp. BGH (Federal Supreme Court), 13 July 2020 (KRB 99/19), NZKartR 2020, 602, 604 et seq.  (2) AEUV, annot. 36; see also text infra at note 82. 47 As regards, e.g. contracts that are not anticompetitive as such, but form part of a network of contracts that amounts to an anticompetitive distribution system see Mestmäcker and Schweitzer (2014), p. 301 et seq. (ibid. also as regards agreements on cooperative joint ventures and rules on concertation in the relationship between the parent companies). In fact, the distinction between implementing agreements and follow-on transactions is artificial insofar as follow-on transactions transpose the anticompetitive agreement into practice and economic reality. generally accepted, and confirmed by the CJEU, 50 that, in principle, abusive contracts, such as exclusivity clauses, non-compete obligations, agreements on price-or quota-fixing or on market division, all whether horizontal or vertical, may be or must be subject to the sanction of nullity, and so must the tying arrangements that Art. 102 lit. d) TFEU outlaws specifically. It is also generally held that this nullity basically corresponds to that of Art. 101(2) TFEU as regards its ordre-public nature, its limitation to the abusive parts of a contract, the issue of severability of such parts from the innocent ones and the fate of these remaining parts, in particular the legal concepts for maintaining them as a ''sound'' transaction. 51 These analogies seem to be all the more justified as, given its restrictive content, the abusive contract typically may also be brought under Art. 101 TFEU 52 so that both prohibitions will be applied simultaneously and, most likely, also in parallel. Such approach, however, risks running against the equally well-recognized principle that the consequences of an infringement of Art. 102 TFEU ought to be determined in view of the specific goals of competition policy underlying that rule of prohibition. 53 Therefore, the obvious difference between Art. 101 and Art. 102 TFEU needs to be taken into account, which is that Art. 101 TFEU concerns agreements whose anticompetitive object or effects are bilaterally, if not mutually, agreed upon between the parties whereas Art. 102 TFEU aims at contracts whose restrictive content typically is unilaterally imposed on the other party by the marketdominating firm. Consequently, particular care must be taken not to punish that party by the nullity of the contract. 54 Rather, the latter's interests in competition and its typically inferior market position may need to be protected by pronouncing a nullity verdict that is appropriate by its being targeted specifically at the abusive exercise of its market power by the market-dominating firm. 55  More particularly, in cases of the concurrent applicability of Arts. 101 and 102 TFEU, the marketdominating firm should not be allowed to invoke the invalidity of the agreement. Such ''asymmetric'' nullity follows directly from the competition-policy rationale of specifically sanctioning the abuse rather than from considerations of an unclean-hands defense, see also supra notes 39, 40. Similarly, if and insofar as follow-on transactions may be held invalid, it may be inappropriate to extend the nullity sanction to transactions entered into by the victim of an abuse. 55 See supra note 50. assessing in practice the many and heterogeneous forms of abuses listed by Art. 102 TFEU or covered by its general concept of abuse of market power. 56 However, in case the abuse consists in a restriction of the freedom of competition of the other party on the relevant market or on upstream or downstream markets, a rule of thumb may be that it is only the other party that may invoke nullity.
2.2.2.2 Concurrent application of Art. 101 and Art. 102 TFEU The differences of purpose, operation and effect of invalidity sanctions regarding agreements infringing Art. 101 or Art. 102 TFEU, respectively, should caution against the concurrent application of both rules of prohibition. 57 Such caution is the more warranted as Art. 101(2) TFEU may be applied too easily and broadly, and be seen as the overriding rule of Union law whereas nullity as a sanction of infringements of Art. 102 TFEU rests ''merely'' on the domestic laws of Member States. Unfortunately, the applicability of national law, i.e. the laws of 27 Member States, presents problems of its own. Given the complexity and the controversies about the proper design and determination of contract invalidity as a sanction of infringements of Art. 102 TFEU under German law 58 -difficulties that are likely to exist under the laws of other Member States as well -the applicable national laws may diverge, the result being that similar abuses are sanctioned differently in the various Member States. The principles of equivalence and of effectiveness of national enforcement of EU law that the CJEU has developed with a view to ensuring the primacy and the full effect of EU law in Member States, and also with a view to making the tasksharing between the Union's legal order and that of Member States work properly and in the common interest of the Union, will not help to solve the problem. Thus, the principle of equivalence accepts national competition law, its sanctions and its enforcement as they are; it only asks for the equal application of the enforcement 56 One example is the controversy regarding predatory pricing; see Fuchs (2019a), Art. 102 AEUV, annot. 429; contra Schröter and Bartl (2014) 58 There is a natural inclination to determine the appropriateness of nullity sanctions for infringements of Art. 102 TFEU by reference to the sanctions provided for or practiced with regard to infringements of similar rules of prohibition of national law (see inter alia Westermann (2020), p. 456 et seq.), and the principle of equivalence (see text at note 59) even invites one to do so. As a consequence, national particularities are transposed to the EU level, a tendency that becomes particularly problematic in cases where the interpretation of the national rule of prohibition is itself controversial and/or the national and the EU rules of prohibition are only seemingly similar. rules and procedures regarding violations of domestic competition law to infringements of the EU's rules on competition. 59 The principle of effectiveness does not ensure uniformity of national laws and procedures of enforcement of EU law, but only requires that such laws and procedures give EU rules of prohibition the full effect that is needed to satisfy their regulatory objectives. 60  principles apply independently, albeit in complementary ways, they cannot substitute harmonization of Member States' laws, nor are they intended to bring about harmonization. 61 Rather, they establish a systemic framework within which Member States preserve their ''enforcement autonomy'' as a matter of their sovereignty and of the split-level structure of the EU's legal order in general, and, more particularly, of its system of implementing its competition policy. Therefore, the development of a concept of contract invalidity under Art. 102 TFEU that specifically addresses abuses of market power and applies throughout the EU in a uniform manner remains an open issue. Such development is all the more needed as, by definition, infringements of Art. 102 TFEU, like those of Art. 101 TFEU, always have an interstate trade dimension, and as, due to the presence of market dominance upon which the abuse rests, they weigh heavily on competition on the relevant markets.

The Consequences of Nullity: Dealing with the Diversity of National Laws
The problem of the diversity of applicable laws takes yet another dimension when it comes to determining the consequences of nullity of agreements pursuant to Art. 101(2) TFEU. Thus, Member States' laws differ as regards their responses to the question whether the invalidity of parts of an agreement that violates public policyrelated rules of prohibition entails the nullity of the entire agreement or not, 62 whether when a cartel has been established in some corporate form the retrospective nature of the sanction of nullity ex lege also results in invalidity with effect ex tunc of an already implemented and practiced organizational structure, 63 whether and how the innocent parts of a transaction may be maintained by re-interpreting the transaction in the light of the hypothetical, objectively understood intentions of the 61 See Weyer (2003), pp. 318, 323 et seq. This does not rule out that, by virtue of their obligation to interpret and apply national law in the light of the objectives of the prohibition rule of Art. 102 TFEU, the CJEU may require Member States' courts to give optimal effect to the sanctions available under national law (see supra note 60). For instance, it may oblige them to apply the sanction of invalidity on their own motion and/or with effect ex tunc, should national law not already so provide. The long-term result of such approach will be some practical approximation of national laws. parties, and whether even anticompetitive elements of a transaction may be saved by reducing their restrictive substance to an acceptable one. 64 National laws also provide for different answers to the issue of whether the enforcement of a ''regular'' business transaction that contains anticompetitive elements may be blocked by a misuse defense of some kind, be it an unclean-hands defense, 65 or be it a defense against a party invoking nullity only as a pretext covering its changed interest in the performance of the transaction, 66 and they diverge on how to treat the follow-on transactions mentioned above. 67 This paper is not the place to enter into an examination of the details of such divergences. Only a few points may be briefly submitted for further discussion.
2.2.3.1 Severability One such point concerns the issue of severability and its treatment with a view to maintaining a transaction on the -frequently implicitassumption that it is basically sound or re-designable into acceptable shape. This is a matter to be dealt with by the national judge according to the law of Member States, in particular their law of commercial contracts. 68 However, the natural inclination of commercial contract lawyers towards severability and maintaining apparently 64 See -also as regards contractual safeguards against full nullity of the entire transaction -Jaeger (2020), Art. 101 (2)  The problem with saving excessively restrictive clauses of such transactions by ''reducing'' them to legality is, first, that the transaction remains one that was not negotiated under conditions of free competition; second, that, as saved, the transaction continues to be restrictive of competition, albeit to a lesser degree; and, third, that by providing for a non-negotiated default mechanism, exit from the restrictive agreement may become less attractive. The approach simply constitutes an invitation to settle on alternative restraints of competition, and, e.g. in the case of no-challenge clauses in license contracts (supra note 42 and infra note 69) it amounts to allowing an indirect ''Kartellzwang'' (indirectly forced cartel adherence workable transactions needs to be contained within strict limits. It amounts to protecting business arrangements that supported an anticompetitive agreement and that allowed it to produce its effects on the market, possibly even for a long time. 69 National law that generously offers such a fallback position is but an invitation to try to profit from such opportunity for practicing unduly restrictive agreements or contract clauses for at least a while. It may also make the judge miss the reality, which is that no such restriction is entered into for nothing, but constitutes part of the bargain or else there is no reason for its being adopted in the first place. 70 There is, therefore, a risk here that national law undercuts the effectiveness of Art. 101(2) TFEU, which, as a sanction, aims not only at parties being free to disregard their anticompetitive agreement as a matter of law, but at ensuring at all times both their freedom to compete and, more importantly, that of the other market actors. 71 This risk is exacerbated by the fact that the parties may choose the application of that law to their agreement or interstate transaction that is most favorable to upholding their transaction despite the invalidity of its anticompetitive components. As necessary as the free choice of applicable law is for international commercial transactions, and as beneficial as it may be that this way only one law will apply to them, in the context of ensuring the uniform effectiveness of EU competition law it hides a particular problem of diversity. This is that the consequences of the nullity sanction of Art. 101(2) TFEU, and thus its actual effect, will vary depending on which of the 27 national laws of Member States will apply, and will so apply possibly by virtue of the parties' choice. This deficit of uniformity will most likely go at the expense of the party to the transaction that is disadvantaged by the 69 This is obvious enough in cases of price, quantity or territorial restrictions, even though part of their effects on competition might be compensated for by the availability of claims for damages. Frequently enough, however, the negative effects of anticompetitive parts of a transaction are less visible and the harm done hard to estimate in terms of monetary compensation. Non-attack clauses in patent licenses are but one example among others, because there is no way of telling whether and how many competitors have been deterred from market entry by the existence of questionable patents the licensee was best placed to challenge successfully. Yet, it is precisely this exclusionary effect (not the equities between the parties to the license) that matters, and it is real (see for the persistent problem of patent quality and the high invalidation rates of litigated patents Ullrich (2020b), passim). This is also why, in principle, Art. 5(1), lit. b) of Commission Reg. 316/2014 (supra note 64) excludes no-challenge clauses from the benefit of a group exemption, and why the ever-recurring criticism of their strict treatment under competition law is misplaced, see Commission, Guidelines for the application of Art. 101 TFEU to technology transfer agreements, OJEU 2014 C 83,3 at paras. 133 et seq., 242 et seq.; A. Fuchs (2019b), Art. 5 TT-GVO, annot. 13 et seq. with references. 70 Once again, no-challenge clauses in license agreements may serve as an illustration of how doubtful it is under competition law to hold them severable from the patent license transaction (as German courts tend to do; see -affirmatively -Schmidt (2019), Art. 101(2) AEUV, annot. 22) and to favor their ex post judicial transformation into a right to terminate the license in case the licensee attacks the validity of the licensed patent: The frequently expressed demand for at least such a right to termination is but evidence of the close economic and contractual link between the non-attack clause and the license transaction. Besides, the general acceptance of a right to termination instead of an outright no-challenge clause has turned out to be questionable in itself; see Art. 5(1), lit. (b) Commission Reg. 316/2014 as compared to Art. 5(1), lit. (c) of the predecessor Commission Reg. 772/2004of 27 April 2004, OJEU 2004 As third parties they would have to bear the risks of a lack of transparency and legal certainty, because they have no way of telling whether transactions by their competitors that come with anticompetitive elements will be severable from the latter or whether such elements might be re-read by a court in ways that make their hypothetically redrafted terms acceptable under competition law. restrictive parts of the business arrangement; it certainly goes at the expense of third parties who cannot tell in advance with certainty whether or not or how much of an impact the transaction will ultimately have on their position in competition. 72 2.2.3.2 Follow-on transactions Another point relates to the issue of whether the nullity of an anticompetitive agreement entails that of follow-on transactions that a party to the agreement concludes on its basis with third parties. The CJEU has referred the matter expressly to Member State law. 73 However, this general referral is questionable to the extent that it is the follow-on transaction that translates the anticompetitive agreement into a market reality, and, thus, puts its distorting effects on competition into operation. 74 Moreover, the repeated conclusion of follow-on transactions may be tantamount to a concerted practice that Art. 101(1) TFEU likewise prohibits. If national law alone applies, and if Member States' laws diverge in respect of the fate to be given to follow-on transactions, as they apparently do, 75 interstate-relevant cartels and other anticompetitive arrangements will likely enjoy more favorable conditions in some Member States than they do in others. Concomitantly, the reach and enforcement of Art. 101(1) TFEU will vary nationally. Thus, in Germany, the still prevailing opinion considers follow-on transactions to be immune against Art. 101(2) TFEU because, as such, the transaction has no anticompetitive object of its own. 76 In addition, extending the invalidity of the anticompetitive agreement to follow-on transactions would create confusion on the market and, possibly, disadvantage the innocent party that has no way of knowing whether or not the transaction is the result of an anticompetitive agreement, but may need protection as regards obtaining performance under the follow-on contract. Critics 77 argue that the CJEU's ruling in Courage v. Crehan 78 has made it clear that under Art. 101 TFEU ''any individual'' 79 is entitled to protection against such harm of anticompetitive agreements as may materialize precisely in -overpriced or otherwise distorted -follow-on transactions. They also maintain that third parties, clients of the parties to the anticompetitive agreement, cannot be expected to abide by the follow-on contract only to thereafter have to bring a tort action for compensation of the damage they suffered. 80 In fact, the issue is not one of determining the ''consequences'' of the nullity of anticompetitive agreements, i.e. whether such nullity necessarily also entails the invalidity of upstream or downstream transactions by the parties to the unlawful agreement. Rather, it is one of determining the reach of the prohibition of Art. 101 TFEU by reference to its purpose. Seen from this perspective, Art. 101(2) TFEU is but the expression of the Treaty makers' objective to annihilate the very operation and effects of the anticompetitive agreement and, by outlawing not only the existence, but also the execution of the agreement, to prevent competition from being distorted. 81 Therefore, the emanations of restrictive agreements, the actual outcomes of its operation in form of distorted transactions, ought to be brought within the reach of the sanctions by which the prohibition may be enforced. This need not be done in an undifferentiated way. Recently, the German Federal Supreme Court has restated its ruling that a notification of termination of a multilateral contract is void if, instead of being decided autonomously by each of the parties, the right to termination is exercised upon coordination -by agreement or otherwisebetween all or several of the parties to the multilateral contract. 82 Arguably, these rulings may be extended to agreements having as their object a restriction of competition whose transposition into follow-on transactions directly results in these being distorted in their very substance. 83 The basis for such extended reach of the 80 Füller (2016), Art. 101 AEUV, annot. 462, while critical of the opinion prevailing in Germany, considers follow-on transactions to be necessarily valid or else tort claims for damages would fail for lack of a legally recognized damage. However, it is precisely the invalidity of the follow-on transaction that constitutes the harm, and in Courage (supra note 78) the CJEU made it clear that a party to an anticompetitive agreement may claim damages despite (or rather because of) the nullity of the agreement. 81 Rightfully stressed by Säcker, supra note 74. according to which a follow-on transaction -the public procurement contract resulting from a bid-rigging agreement -is valid to the extent that validity serves to protect the third party's interest in performance of the contract. 83 Limiting invalidity to follow-on transactions as defined here, i.e. follow-on transactions between a party to the agreement and a third party (first-level follow-on transactions) serves easy identification of the relevant follow-on transactions. Such limitation results from an analogy to transactions constituting or implementing a concerted practice; it ensures the existence of a link of causality between the anticompetitive agreement and the distortion of the substance of the follow-on transaction. By contrast, the prevailing opinion would remedy such distortion only if the third party was actually deceived about the existence of an anticompetitive agreement when entering into the transaction, mere ignorance of such agreement being irrelevant under contract law. Most often, however, it is the latter hypothesis that corresponds to reality since cartels are clandestine by nature. invalidity rule of Art. 101(2) TFEU is that Art. 101 TFEU establishes an ordrepublic prohibition that encompasses all acts of infringement, such as practicing the anticompetitive agreement by concluding follow-on transactions. 84 This approach has much to offer: It enhances uniformity of the effects of the sanction throughout the EU on the one hand, and, on the other, its legitimacy in that it activates the interests of third parties that are directly affected by the anticompetitive agreement, often enough as regards their own competitive position. If Art. 101(2) TFEU is understood and applied according to its purpose, challenging the validity of the follow-on transactions may help them to effectively protect their interests. 85 In addition, it is a necessary option, because the availability of relief by tort actions for damages provides no adequate substitute sanction, 86 but only a complement.

Compensation
There is, of course, no denying the importance of sanctioning infringements of the Union's rules on competition by the grant of claims under tort law. However, the purpose of this paper requires limiting the presentation of the availability of relief from such infringements by the law of torts to the role such claims for compensation may play in comparison to the sanction of nullity of anticompetitive agreements. To this end, a short summary of the principles of tort liability for violations of Arts. 101 and 102 TFEU and of the practical operation and importance of the enforcement of such liability may suffice.  (KZR 7/17), NZKartR 2020,196,197, para.30 -''Einspeiseentgelt III(TV cable transmission fee III)''. However, this idea will produce counterproductive results if third parties ''protected'' in this way are held to follow-on transactions that clearly are disadvantageous to them, as is the case of follow-on transactions to price or quota cartels. An obligation to pay the cartel price as a quid pro quo for obtaining legal certainty can hardly constitute a reasonable result of competition policy. Instead of stressing third parties' interests in legal certainty, a differentiating approach is needed. Thus, third parties may be given the option to claim invalidity or to take the transaction as it is, e.g. in cases of rebate cartels the third party may wish to benefit from the reduced price. Such differentiation would follow directly from the purpose of the nullity sanction rather than being dependent on contract law considerations regarding willful deceit, misuse of rights or fairness in commercial dealings, which latter are misplaced in the context of sanctioning anticompetitive practices anyway. 86 Contra Bechtold (2020), p. 461; Schmidt (2011), pp. 572 et seq., 576. 101(2) TFEU), liability may exist only if and to the extent Member States' laws provide for it. However, pursuant to the CJEU's ruling in Courage v. Crehan and its subsequent case law, 87 the combined operation of the direct effect of Arts. 101 and 102 TFEU and of the principle of effectiveness have as a result that violations of the prohibition of anticompetitive agreements, of concerted practices and of abuses of market dominance must, as a matter of Union law, be qualified as tortious acts that are individually actionable under national law, with the principles of equivalence and effectiveness conditioning the nature, scope and enforcement of protection. 88 As regards German law, this means that relief must be granted from the infringement by way of injunctions ordering the infringer to cease and desist from its unlawful conduct and by the award of compensation for the harm done. 89 While the grant of injunctive relief, although important as regards many forms of infringement of Art. 101 TFEU, and in particular of Art. 102 TFEU 90 (and also of related rules of prohibition of national competition law 91 ), has not attracted the attention of the EU legislature and has been brought to the attention of the CJEU only rarely, 92 Member States rules on the substance of and the procedure for enforcing claims to compensation have done so repeatedly. Thus, the CJEU's ruling that ''any individual'' may claim compensation where there is a causal relationship between the damages suffered and an anticompetitive practice 93 extends protection 87 See supra note 60. As regards tort claims, it extended their availability from claimants the law is intended to protect specifically, as provided for by the general rule of § 823(2) BGB (German Civil Code), to any individual that has been harmed, see Explanatory Notes to § 33, WuW 2005, ibid. at p. 167 et seq. Note that § § 33, 33a GWB differ as regards entitlement to injunctive relief ( § 33) and entitlement to a claim for recovery of damages; only the latter claims benefit from the broader definition of entitlement provided for by Dir. 2014/104, see infra note 98 and Roth (2019) not only to indirect damages suffered by market actors on the demand side of downstream markets, 94 in particular by end-consumers, with the ensuing need for providing for some form of collective enforcement. 95 Rather, it also brings within the scope of an infringer's liability damages that market actors suffer as a result of high prices that outsiders to the anticompetitive arrangement can set in the shadow (or ''under the umbrella'') of a cartel, 96 and it even allows non-market actors to recover losses incurred indirectly due to an anticompetitive practice or arrangement. 97 By its Arts. 1, 3 and 4, Directive 2014/104/EU of 26 November 2014 98 adopts the CJEU's Courage/Crehan and Manfredi/Lloyd Adriatico rulings and also the principles of equivalence and effectiveness 99 as the basis for harmonizing Member States' competition laws in regard of various elements of tort liability. These are the scope of recoverable damages (Art. 3(2),(3)), 100 recovery of overcharges by indirect purchasers and the passing-on defense including avoidance of over-compensation and the burden of proof (Arts. 12 to 14), 101 accommodation for actions by multiple claimants from different levels of supply chains (Art. 15), and joint and several liability of multiple infringers (Art. 11). 102 As a matter of no lesser practical importance, Directive 2014/104 obliges Member States to facilitate actions for 94 The concern generally is with harm done to actors on downstream markets. Dir. 2014/104 (infra note 98) as well seems to cover only such hypotheses (see Arts. 13,14). However, since the exercise of coordinated or of monopolistic buyer power may cause harm on upstream markets, direct and indirect suppliers of the infringer must be entitled to recovery of damages as well; see Roth (2019) damages by specific rules on disclosure of evidence by defendants or third parties (Arts. 5,8) or by a competition authority (Arts. 6, 7), 103 by rules on the quantification of harm, including the possibility to estimate the amount of damages suffered, and by a -rebuttable -presumption regarding the existence of a causal link between the harm suffered and the alleged anticompetitive practice (Art. 17). Finally, Art. 10 requires Member States to introduce limitation periods of at least five years, and specifies their practical terms. 104 2.3.2 Problems: Popular, but Only ex post By harmonizing and strengthening national law on actions for damages at its ''critical'' points -scope and calculation of damages, passing-on defenses, access to evidence and burden of proof -the Damages Directive has given support and breadth to a trend towards tort-law based private enforcement that has its roots in both the CJEU's Courage/Crehan case law 105 and the shift from the centralized administrative enforcement system of Reg. 17/62 to the decentralized system of Reg. 1/2003 whereunder national competition authorities and courts are called upon to apply directly both Art. 101 (1) and (3) TFEU. 106 To the general satisfaction of the EU legislature, 107 actions claiming damages under national law for infringement of the EU's rules on competition have become almost synonymous with their private enforcement. Litigation rates have risen to unprecedented numbers and, due to the amount of damages claimed and sometimes obtained, 108 attracted public attention. However, when looking at the -admittedly incomplete 109 -statistics, 110 doubts arise as to the systemic effects of such tort-law based private enforcement. In only 2% of the 239 litigation cases brought since 1998 were the actions brought on a stand-alone basis whereas 98% were brought only once an administrative decision had been taken by the Commission (40%) or by national authorities (57%) confirming an infringement of competition law. The reason is that, as regards the ascertainment of a violation of Art. 101 or 102 TFEU, these so-called ''follow-on actions'' benefit from the binding effect of such administrative decisions. 111 About half of the actions originate in the public sector, apparently concerning bid-rigging cartels, the other half in the private sector, apparently all concerning hardcore (price) cartels; only very few of the actions were brought by end-consumers.
These figures suggest that private enforcement by tort action for damages does not represent an independent way of enforcement of the Union's competition rules or broaden its scope, but only comes on top of administrative enforcement in that it adds damage claims to public fines. As such, it comes belatedly once the harm is 109 Private enforcement based on invalidity claims (or defenses) or on claims for injunctive relief apparently has not been examined statistically in recent years; for an early study Peyer (2012) (2019), p. 136 et seq. Note that Laborde bases his statistical analyses on counting cases rather than judgments, i.e., as a rule he counts by judgments, but not when several legally similar judgments have been rendered in respect of the same base litigation. Thus, the 239 cases represent 368 judgments. By contrast, Sailer (2018), passim, counts follow-on litigation by reference to cartels as prosecuted by the Commission and national competitions authorities. This gives an interesting overview of the existence of litigation in EU Member States regarding a considerable number of cartels but inflates the number and the scope of litigation. Thus, in Germany alone the truck cartel (Commission, decision of 19 July 2016, OJEU 2017 C 168,6) has triggered more than 400 follow-on actions for damages or 80% of all follow-on actions. By contrast, 20 decisions of the European Commission or of the Federal Cartel Office triggered ''only'' 50 follow-on actions; see Bundeskartellamt (2019), p. 23; Bundeskartellamt (2017Bundeskartellamt ( /2018. Note also that the number of out-of-court settlements is unknown as is the volume of settlement by way of arbitration, both presumably covering mainly the determination of damages once basic liability has been confirmed judicially; see Bundeskartellamt (2019) . The legitimacy and/or the scope of the binding effect may also be doubted if a decision on fines is the result of a settlement procedure, a way that is frequently taken, but remains unclear as regards its boundaries, see only Barbier de la Serre and Lagathu (2019) done, and frequently only when the anticompetitive arrangement is about to break down anyway. 112 In addition, it reinforces an already existing enforcement asymmetry, which is the focus of much administrative enforcement effort on prosecuting outright violations of the competition rules by cartels, i.e. serious infringements, which, however, raise more problems of fact-finding and fact assessment than issues of normative evaluation.
It is, indeed, also mainly such hardcore cartels that produce measurable overcharge effects, which then translate into individual harm in the form of quantifiable damages suffered. There is, of course, no question that such damages need to be compensated for. However, at the time this is actually done, 113 the conditions of competition and/or the business interests may have changed, and so may have the management responsible for the infringement. The damages due will have become a matter of making appropriate, possibly even burdensome reserves on the balance sheet, but the strategic and/or tactical objectives of the anticompetitive arrangement or practice have been attained. 114 More particularly, supra-competitive prices have been obtained 115 and market presence is maintained, if not reinforced at the expense of competitors.
112 For much of their prosecution activity competition authorities (have to) rely on so-called leniency programs under which a participant to a secret cartel may remain unpunished or obtain a reduction of its fines as a ''reward'' for cooperating with the competition authority by disclosing information on the creation and operation of the cartel; see Art. 2, Nos. 15, 16 Dir. 2004/104; Commission, Notice on immunity from fines and reduction of fines in cartel cases (OJEU 2006C 298, 1, as amended OJEU 2015. For Germany see § § 81h et seq; here, the number of such disclosures for benefitting from leniency saw a peak of 76 requests in 2015, but has been declining ever since with only 16 requests in 2019; see Bundeskartellamt (2019), p. 34; Bundeskartellamt (2016), p. 20. In addition, these ''immunity recipients'' benefit from a limitation of their joint and several liability under Art. 11(4), (5), (6) Dir. 2014/104. Whether the decreasing number of requests will result in lower numbers of tort actions for damages remains to be seen. The point here is that such requests will be made only once a participant of a cartel anticipates the detection of the cartel or wishes to exit the cartel for business reasons anyway, i.e. always when the cartel is about to break down because there is no reason for a party to a cartel to forego its benefits before this. 113 According to Laborde (2019), p. 7, No. 39, the infringement decision on average came 8.4 years after the first purchases had been made at cartel prices, and the first civil judgment was rendered 4.2 years later; for Germany Rengier (2018), p. 618 et seq. finds that on average first instance court decisions take 1 years, judgments on appeal another 20 months, and judgment on second appeal on grounds of law yet another year, which adds up to more than four years of court proceedings or more than three years in reality because review by the Federal Supreme Court is not generally sought. 114 According to Bundeskartellamt (2019), p. 23 settlement on the amount and payment of damages may in part result in compensation through price adjustments in future periods of supply, i.e. the supply relationship is maintained. 115 Supra-competitive prices include the not unlikely hypothesis of participating firms operating at noncompetitive costs and/or very low profit margins, which, ultimately, may mean that claims for damages, however well founded, will be an ineffective because fruitless sanction. In extreme cases, even fines may have to be set so as not to lead to full failure of the firm; see Commission, Guidelines on the method of setting fines pursuant to Art. 23 (2),(a) Reg. 1/2003, OJEU 2006Nowak (2020), Art. 23 VerfO, annot. 36 (p. 1410).

Conclusion
Clearly, as a sanction, invalidity has its limitations and drawbacks, too. It does not stop parties from actually practicing coordination in competition, and, generally speaking, it is better suited to sanction horizontal restraints of competition than vertical ones. However, it hurts the parties to an anticompetitive arrangement where they are most vulnerable, namely directly on the market that they wish to shield or capture. It is this idea that, by being entirely free to exit the anticompetitive agreement, parties may at any time assess afresh their interests and return to undistorted competition and that, as a consequence, the effectiveness and the reliability of the arrangement will suffer sufficiently to make anticompetitive agreements unattractive. Therefore, this idea needs to be buttressed as best as possible rather than have it contained by so many considerations of contract law limiting invalidity with a view to saving the transactions inter partes. Competition law is primarily concerned with the effects anticompetitive agreements produce erga omnes in regard of third parties. 116 It is, of course, sound competition policy to differentiate when applying the nullity sanction so as to penalize the ''true'' infringers and not also their ''victims''. In the absence of a specific rule that in cases where invalidity typically produces asymmetric or even perverse effects would make invalidity dependent on being claimed by the party victim of an anticompetitive agreement and unavailable for the other party, 117 such differentiation needs to be developed by applying Art. 101(2) TFEU or a nullity sanction for abuse of market dominance (Art. 102 TFEU) under national law from the perspective of competition law and with a view to fully attaining its goals. This will require systematic category-building with a view to classifying the many different types of restrictive agreements according to their exit conditions and consequences, and it will require a doctrinal willingness to complement concepts of contract law by considerations of competition law. The purpose of this paper was more limited, namely to raise awareness of the trend towards identifying private enforcement with tort-law-based claims for compensation of harm and, thus, toward neglecting invalidity. After all, enforcing invalidity is genuinely private in that the initiative for and the aim of pursuing it require more than a basically defensive reaction to harm suffered -often commonly -due to other market actors' unlawful conduct, namely a pro-active determination of the individual firm or of consumers to autonomously 116 This holds true also for vertical restraints such as agreed upon as part of the establishment of distribution systems or supply chains because they affect third parties' access to distribution or supply and/or consumer choice. Likewise, non-competition clauses are not only a concern as regards competition between the parties, but mainly as regards competition with other market actors and/or third parties' access to alternative sources of supply or to alternative outlets. 117 Van Gerven (2003), European Competition Law Annual 2001, at 57 seems to suggest the applicability of Art. 6 of Directive 93/13 on Unfair Terms in Consumer Contracts (OJEC 1993, L 95, 29, as amended by Directive 2011/83/EC, OJEC 2011, which obliges Member States to provide in their national laws that unfair terms in consumer contracts ''shall … not be binding on the consumer and that the contract shall continue to bind the parties upon those terms if it is capable of continuing in existence without the unfair terms''. However, in contrast to unfair terms in consumer contracts, anticompetitive clauses of business transactions typically form part of a negotiated deal, and there are hardly ever default rules of general contract law that could fill the gap, if any, left by the invalid clause. assess and act in conformity with its or with their particular contractual and competitive positions. Moreover, private enforcement through challenging the validity of agreements in restraint of competition is independent of administrative enforcement. In fact, if the invalidity attack is brought only once competition authorities have initiated prosecution, it will come too late.
Funding Open Access funding enabled and organized by Projekt DEAL.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creativecommons.org/licenses/by/4.0/.