The Principle of Effectiveness, Competition Law Remedies and the Limits of Adjudication

The principle of effectiveness is closely related to the development of the emerging EU law on remedies. Its instrumental use has enabled the EU courts to restrict the principle of national procedural autonomy, when this was convenient in order to ensure the accomplishment of the aims set by EU competition law enforcement, and to establish EU-granted remedies, the most notable one being the right to claim competition law damages. The principle of effectiveness may also influence the design of injunctive relief by the European Commission, which should be adequate to ensure not only that the results of the violation of competition law are reversed, but also that there is no risk that the aims of competition law will be jeopardized in the future (general deterrence, specific deterrence and prophylactic/preventive aims). Left unbound, the principle of effectiveness may offer the opportunity to competition authorities to expand their remedial discretion and to re-design market processes and outcomes in accordance with the dominant interpretation of their statutory objectives. The point made in this paper is that, whatever one thinks of the appropriateness of an expansive view of remedial discretion, which is not, in our view, supported by the restrictive interpretation of the principle of effectiveness in EU law, remedial discretion is naturally limited by the specific function exercised by the remedial process chosen or, more contentiously, imposed by the nature of the dispute. Drawing on Fuller’s account of the existence of various forms of social ordering, each of them emerging in specific circumstances/context and having its own principles and limitations, the paper offers some reflections on the possible limits that the essence of each ideal type of social ordering sets to the expansive interpretative potential of the principle of remedial effectiveness. The polycentric nature of competition law disputes calls for flexibility in the choice of the adequate form of social ordering aiming to achieve the objectives set by the legislator. This breaks with the classic view of the adjudication model and hints to the prevalence, in a significant number of cases with a pronounced polycentric element, of what has been called the “structural adjudication�? model, still distinct from the model of regulatory governance. The paper explores the nature of commitment decisions as an illustration of the difficulties of classification, without a proper consideration of the functions and respective limits of each form of social ordering. It does this by examining some recent cases, such as the ongoing Google saga at the European Commission or the Skyscanner judgment of the UK Competition Appeal Tribunal (CAT).


The Principle of Effectiveness, Competition Law Remedies and the Limits of Adjudication
Ioannis Lianos 1

I. Introduction
The principle of effectiveness is closely related to the development of the emerging EU law on remedies. Its instrumental use has enabled the EU courts to restrict the principle of national procedural autonomy, when this was convenient in order to ensure the accomplishment of the aims set by EU competition law enforcement, and to establish EU-granted remedies, the most notable one being the right to claim competition law damages. The principle of effectiveness may also influence the design of injunctive relief by the European Commission, which should be adequate to ensure not only that the results of the violation of competition law are reversed, but also that there is no risk that the aims of the competition law enforced will be jeopardized in the future (general deterrence, specific deterrence and prophylactic/preventive aims) 2 . The principle of effectiveness may therefore reinforce the remedial discretion of the Commission and national competition authorities when implementing EU competition law. Discretion to adopt effective remedies cannot, however, be unlimited.
There are various forces that may operate in order to keep it in check. First, due process rights and fundamental rights jurisprudence, following the conferral of a binding effect to the EU Charter of Fundamental Rights and the expected integration of the EU in the ECHR system may balance the need for effective, read expansive, remedies and therefore limit the risk of discretionary remedialism 3 . Second, wellestablished conceptions in private and public law about the inherent limits of the remedial exercise may also exercise constraints to the remedial discretion of the Commission and other competition authorities: the principle of proportionality, for public law, or the correlativity of private law disputes cannot accommodate fully the demands of optimal enforcement theory for deterrence. For instance, remedies should fit the competition law problem identified at the liability stage and the theory of harm the decision-maker relied upon in order to find the defendant liable for the competition law infringement 4 .  3 -Discretionary remedialism‖ is the view that courts and competition authorities have discretion to award the -appropriate‖ or -optimal‖ remedy in the circumstances of each individual case rather than being limited to specific (perhaps historically determined) remedies for each category of causative events. What is optimal is in these cases decided according to the interests and aims of the decisionmaker. 4 For analysis of these limits to discretionary remedialism and the need for remedial discretion to be compatible with established legal conceptions of remedies, see I. Lianos, Competition law remedies in This chapter explores a further limit to remedial discretion, that of the legitimacy of decision-makers. Legitimacy in this context does not only relate to the fact that public action should be legitimate from a legal perspective, that is, not illegal. Because of its potential for generating expansive remedies, the principle of effectiveness may render this purely legal limit ineffectual. Legitimacy, in this context, refers to the specific legitimacy-building mechanisms elaborated by the specific social order in order to guarantee that its action is politically acceptable by those to which a remedy will be imposed. As different social order systems dispose of different mechanisms to guarantee the legitimacy of their action, it is important to guarantee that the use of the powers conferred to each system corresponds to the specific functions exercised by it and do not expand to the realm/functions exercised by other forms of social ordering, which dispose their own legitimacy-generating mechanisms. If such trespass occurs, it may create a legitimacy crisis. In the event of a trespass, the decision-maker should therefore emulate the legitimacy-building tools of the social order whose boundaries were crossed in order to guarantee the legitimacy of its action. More concretely, it is argued in this Chapter that remedial discretion is an instance of adjudication and, as such, it is subject to the inherent limits of adjudication, as a separate form of social ordering from those of contracts/negotiation, managerial/administrative discretion or legislation 5 . Should remedial discretion move beyond the limits of adjudication and cross over the -territory‖ of regulation or contractual governance, the decision-maker should adapt, by developing legitimacy-building tools that would emulate those used in the context of these other forms of social ordering. The participation in the remedial process of the interests affected constitutes an important source of legitimacy for managerial/administrative discretion 6 . Consequently, if competition law remedies move closer to regulatory ones and therefore cross the limits of adjudication, they should give rise to increased participatory rights of the interests affected, including interests others than the parties to the dispute.
I explore the interaction between the principle of effectiveness and remedial discretion, before exploring how the limits of adjudication may be crossed over by the remedial action of competition authorities, in view of the polycentric dimension of competition law disputes. I then criticize in the following section the narrow view of adjudication and advance a distinction between the dispute resolution adjudication model and the structural adjudication model, the latter being closer to the managerial or administrative discretion form of social ordering and thus conducive to resolving polycentric disputes. The final part reflects on the emergence of the -consensual‖ model of competition law enforcement, with the development of the practice of commitment decisions. I argue that, contrary to what has been thought so far, commitment decisions do not constitute a contractual form of social ordering but present characteristics similar to the structural adjudication model. In addition to being theoretically a more accurate description of the reality of such remedial practice, the re-conceptualization of commitment decisions as a form of structural adjudication should enable the integration of legitimacy-building mechanisms emulating those of the managerial or administrative discretion model, such as increased participation of all interests affected by the remedy in the decision-making process.

II. The principle of effectiveness and remedial discretion
For a long time the topic of remedies has been the focus of EU constitutional scholars researching on national procedural autonomy and its limitations by the principles of effectiveness and equivalence. In particular the principle of effectiveness has been expansively interpreted by the European judiciary to encompass all forms of remedial action for infringements of EU law, and was constitutionalized with the adoption of Article 47(1) of the Charter of Fundamental Rights in the EU (hereinafter: Charter) and Article 19(1) sentence 2 TEU 7 . This has led to the development of the emerging field of EU remedies law and has had important implications in the enforcement of EU competition law. Norbert Reich has described the development of the principle of effectiveness in the case law of the CJEU as following three distinctive approaches 8 :  The principle of effectiveness has initially been understood as an -elimination rule‖, enabling national courts to put aside or dis-apply national provisions making the implementation of EU rights -practically impossible‖ or -excessively difficult‖ or to interpret the national remedy in conformity with EU Law. The jurisprudence of the European Courts requires Member States to establish a system of legal remedies and procedures which ensure respect for the right to effective protection of individuals' rights under EU law, and hence put in place effective remedies for violations of EU law. When adjudicating the protection of EU rights in private legal disputes, the CJEU has mostly relied on its ‗effectiveness and equivalence' test to assess national procedures, -effectiveness‖ referring to the requirement that national procedural rules should not make the enforcement of EU rights impossible or excessively difficult..
 The second step in the development of the principle of effectiveness is its -use as ‗hermeneutical' i.e. interpretative principle‖. This provides some positive content to the principle of effectiveness, requiring Member States to provide for remedies -sufficient‖ to ensure the protection of EU granted rights. This involves a twosteps approach from national courts: first to spell out the EU right which must be protected by an adequate remedy; second, the development of a minimum core of what consists an effective or -sufficient‖ remedy to ensure the protection of the EU granted right. it follows from the principle of effectiveness and the right of any individual to seek compensation for loss caused by a contract or by conduct liable to restrict or distort competition that injured persons must be able to seek compensation not only for actual loss (damnum emergens) but also for loss of profit (lucrum cessans) plus interest‖ 9 .
As it transpires from the above excerpt, the CJEU first refers to the principle of effectiveness (-effet utile") in order to recognize an EU-granted right for any individual harmed by a competition law infringement to claim damages for the loss caused by the infringement (the EU right of compensation), before elaborating on the corresponding remedy for that right, the award of damages, which should at least ensure (the minimum core) that the injured person may receive compensation for the abovementioned heads of damages. In essence, the procedural dimension, which, in theory, is left to the national level, is subject to the interpretation of the EU law granted right, and its scope, as this is determined by the CJEU at the EU level. There have been attempts to enlarge this minimum core with additional requirements to be imposed to the national level as to the development of adequate remedies. For instance, in his opinion in Donau Chemie, Advocate General Jääskinen examined the scope of Article 19(1) TFEU advancing the view that -in the light of that Treaty provision, the standard of effective judicial protection for EU based rights seems to be more demanding than the classical formula [of the ‗effectiveness' principle] referring to practical impossibility or excessive difficulty. In my opinion, this means that national remedies must be accessible, prompt, and reasonably cost effective‖ 10 .
The CJEU did not follow the AG's approach on this issue and retained its previous definition of the principle of effectiveness. In VEBIC and in Kone, the CJEU employed the principle of effectiveness (or full effectiveness) in order to derive from the interpretation of Article 101 TFEU an obligation for Member States to entitle national competition authorities to participate, as a defendant or respondent, in proceedings before a national court which challenge a decision that the authority itself has taken (in VEBIC) or the right for victims of umbrella pricing to claim compensation for harm suffered by an anticompetitive conduct in case there is a direct causal link between that conduct and the harm they suffered (in Kone). What is interesting in both cases is the insistence that the aims followed by Article 101 TFEU should not be put -at risk‖ 11 . According to the Court, Those rules must therefore specifically take into account the objective pursued by Article 101 TFEU, which aims to guarantee effective and undistorted competition in the internal market, and accordingly, prices set on the basis of free competition‖ 12 .
The consideration of the objective pursued by Articles 101 (and 102) TFEU seems therefore a passage obligé for the hermeneutical endeavours of the CJEU with regard to the principle of effectiveness. National remedial provisions should be interpreted accordingly so as to avoid any possible risk as to the full realization of the objectives of the abovementioned provisions of the Treaty. One may note the almost unlimited hermeneutical potential of this construction, as to the intervention of the CJEU in curtailing national procedural autonomy.
 The third level of development of the effectiveness principle follows naturally from the previous one and, according to Reich, consists in the development of -hybrid‖ EU remedies, whose function is to -upgrade‖ national remedies. Reich explains that Article 19(1) sentence 2 TEU establishes a clear link between -effective protection‖ and -sufficient remedies‖, thus positing a unidirectional relation between EU granted rights and national remedies: the recognition of EU rights should lead to the development of -sufficient‖ national remedies to fully ensure their respective scope of protection. The absence of -sufficient‖ national remedial tools cannot nevertheless question the scope of the EU granted right. National law should be interpreted creatively so as to ensure the minimum core of effective enforcement for the EU granted right. This requires, according to Reich, a three-steps approach: (i) -the first step is concerned with finding appropriate national remedies in case of violations of EU granted rights‖; (ii) the -national remedy has to be measured against the yardstick of the negative, eliminatory EU principles of effectiveness and equivalence and if the result is unsatisfactory from an EU law point of view as interpreted as providing ‗insufficient remedies', it has to be ‗upgraded' to meet EU standards‖; (iii) -the remedy thus found is a ‗hybrid' insofar as it takes up elements of national law in the limits of the principle of procedural autonomy, as well as effectiveness in its different functions‖ 13 .
Precisely because of this -hybrid‖ character, the definition of what constitutes an -effective‖ remedy is a highly contextual issue, which depends on the perceived adequacy of the national remedial tools at the disposal of EU law and of the scope of the rights and corresponding duties generated by the provision of EU law interpreted. Although the EU courts do not explicitly posit the link between this secondary right (remedy) and a primary right, this is essentially required, the first step in their reasoning consisting in interpreting the substantive provisions of the Treaty in competition law (in particular Article 101 TFEU) so as to derive a primary right -of any individual to seek compensation for loss caused by a contract or by conduct liable to restrict competition‖. By doing so, the EU law goes as far as defining the scope of the secondary right, which should presumably reflect the scope of the primary right it aims to preserve: the ability of the injured person to seek compensation for actual loss, loss of profit plus interest. One may advance that the scope of the secondary right would have been different, presumably wider, if it reflected a different primary right, for instance that of deterring future antitrust violations by the specific defendant (specific deterrence) or that of deterring future antitrust violations by any competition law infringer (general deterrence) 15 . In Manfredi, the Court did not exclude such interpretation of Article 101 TFEU of the Treaty, recognizing national legal systems the possibility to award punitive damages 16 . Yet, it did not interpret Article 101 TFEU as requiring the inclusion of punitive damages in the minimum core of the -hybrid‖ EU remedy of claiming damages for competition law infringements, as this did not relate to the scope of the primary right recognized by the Court (that of receiving compensation) 17 .
In other words, the main function of the principle of effectiveness is to maximise the -sufficient‖ attainment of the ends pursued by the primary right by providing the -adequate‖ content to the secondary right of claiming antitrust damages, and not to sculpt the essence of the primary right.
The largely hermeneutic function of the principle of effectiveness has not escaped the attention of legal commentators, some, for instance, arguing that the principle has been used without much consistency, as a -kind of jack-in-the-box instrument‖ allowing the EU Court to justify almost every result 18 . The -principle‖ of effectiveness was derided as being essentially -an empirical concept‖, requiring the EU courts to proceed to a comparative examination of the national remedies available and also as providing the authorities the discretion to establish a low or a high level of ( enforcement, the principle being an instrument to achieve undefined policy objectives 19 . Although there is some grain of truth on this characterization of the -principle of effectiveness‖, and the criticisms over an inconsistent application of that principle may be spot on, it remains, however, that the principle does not have a substantive content of its own, but reflects the substantive (primary) right to which it is associated. The principle of effectiveness cannot be seen functioning outside the context of the adjudication of the primary right violated by the competition law infringer.
Claims for damages are usually perceived as an archetypical private law dispute, by which I mean a dispute between two parties, whose rights and duties are correlative to each other. For instance, the infringer's wrong consists in the violation of the primary right of the victim. The secondary right of the victim for compensation correlates to the wrong committed by the infringer. Yet, not all competition law disputes may enter in this mould. Some remedies may have a considerable impact on other economic actors than the parties to the dispute, in a direct or an indirect way. Damages claims may also emanate from complex causation chains. One may think of claims introduced by indirect purchasers, counterfactual customers or umbrella customers. It may not always be easy to interpret the substantive provisions of EU competition law as giving rise to specific primary rights, like that of compensation for losses incurred because of the competition law violation, giving rise to secondary rights, such as that to claim damages. It may be, for instance, quite complex to transpose in rights' (primary or secondary), and corresponding duties', talk the infringement of Article 102 TFEU by a loyalty rebate of a dominant undertaking giving rise to a remedy of mandatory injunction. How would the effectiveness of the remedy be judged in this context? Should the primary right be in this case that of the equal opportunity of the dominant firm's competitors to access the market? But, what size of the market would it be considered -sufficient‖ so as to preserve the essence of that -primary‖ right? Would the primary right be that of the consumers' of the dominant undertaking who may be charged higher prices in the future, should the competitors be excluded from the market? In this case the primary right will need to reflect the economic theory of consumer harm advanced, a quite difficult task. Or is the primary right that of the -public at large‖ been deprived from a moral right to the protection of -competition as such‖ 20 ? This will inevitably raise the issue of the delimitation of this primary moral right. The conceptual plasticity of the substantive provisions of EU competition law may lead, if combined with an instrumental view of the principle of effectiveness, to a quite broad remedial discretion for decision-makers.
Similarly, the secondary right (remedy) may be interpreted broadly so as to enable the adoption of competition law remedies that have a prophylactic (preventive) aim 21 . These remedies seek to ensure that there remain no practices likely to result in distortions of competition and infringements in the future. Prophylactic remedies may be distinguished from specific deterrence as they affect the ability (and not the incentive) of the infringers to commit equivalent anti-competitive practices in the future by focusing on specific facilitators of potential infringements. These are not illegal practices in themselves, but in the specific circumstances of the case, they may facilitate illegal conduct. By prohibiting these practices, the decision-maker's objective is not to deter the potential infringers from adopting such conduct, as this is not illegal, but to reduce their ability to commit illegal practices.
In addition, in an economically oriented competition law, the definition of what is -optimal‖, -sufficient‖ or -appropriate‖ remedy may also be influenced by the view economists take on optimal deterrence (the optimal deterrence model) and on how the market equilibrium existing prior to the competition wrong may be improved by subsequent remedial action. The remedy may thus offer the opportunity to design a new market equilibrium, more competitive than the one following the infringement, but also, in some circumstances, more competitive than the one existing prior to the infringement in order to maximize welfare. Linking the principle of effectiveness with the concept of efficiency, perceived as wealth maximization, may nevertheless open Pandora's Box with regard to the remedial discretion enjoyed by competition authorities and the courts. These may adopt far-reaching remedies which would not correlate with the wrong committed and thus the scope of the primary right that was violated.
Yet, -discretionary remedialism‖ cannot be the rule of the game 22 . This is clear in view of the importance the principle of proportionality has in delimiting the remedial discretion of competition authorities and courts in EU competition law, if one takes a public law perspective on remedies. The principle of proportionality requires that measures adopted by Community institutions do not exceed the limits of what is appropriate and necessary in order to attain the legitimate objectives pursued by the legislation in question; when there is a choice between several appropriate measures, recourse must be had to the least onerous one, and the disadvantages caused must not be disproportionate to the aims pursued. This trade-of exercise involves in addition to considering if the means chosen are indeed a rational means to a purported end (step 1 of the test), some assessment of the possible excessive costs of the remedial action in relation to its benefits (step 2) and whether the means chosen are the least restrictive to the affected interests' alternative (step 3). The first step of the proportionality principle is of particular interest for our purposes. The importance of remedial fit is often stressed by antitrust law literature 23 . It is also indirectly linked with the existence of a causal relation between the undertaking's conduct and the theory of harm advanced. The scope of the obligations imposed on the undertakings concerned in order to bring the infringements to an end identified should be implemented according to the nature of the infringement declared and the obligations imposed ‗must not exceed what is appropriate and necessary to attain the objective sought, namely re-establishment of compliance with the rules infringed'. 24 This relates to the obligation of the Commission to give the undertakings concerned the opportunity of being heard on the matters to which it has taken objection. For example, the Commission is not entitled to impose a fine on an undertaking without having previously informed it in the statement of objections that it intended to do so, a requirement which applies by analogy also to remedies. 25 The existence of a competition law violation (even if there is no explicit finding of an infringement) -constitutes the basis of the obligation of the parties to terminate the infringement‖, hence the reason for imposing remedies. 26 However, if the principle of proportionality requires a close fit between the harm and the remedy, determining the nature of that fit (functional, instrumental) remains an open question. The flexibility of the concept of the theory of harm, in particular its linkage to economic theory, enables the Commission or the competition authorities to enjoy a wide remedial discretion, in particular when adopting prophylactic remedies.
While in phase with the principle of effectiveness, prophylactic remedies have nevertheless be considered in certain circumstances as extending unreasonably the scope of the remedy, with regard to the scope of the primary right violated and thus the wrong committed. For instance, in a number of cases, the Commission required the undertakings concerned under its infringement decision to refrain in the future from any conduct which may have a same or similar effect to those covered by the infringement decision, with the aim of preventing the undertakings from repeating the behaviour found to be unlawful. 27 In Cartonboard, the Commission prohibited any future exchange of commercial information by which the participants directly or indirectly obtained commercial information on competitors, even if this was not by its nature unlawful under Article 101 TFEU as the information related also to certain aggregated statistical data. 28 The General Court found that such prohibition exceeded what was necessary in order to bring the conduct in question into line with what was lawful because it was seeking to prevent the exchange of purely statistical information which was not in, or capable of being put into, the form of individual information and thus used for anti-competitive purposes. Indeed, the Commission had not considered the exchange of statistical data to be in itself an infringement of Article 101(1) TFEU. According to the Court, ‗the mere fact that a system for the exchange of statistical information might be used for 24  In Atlantic Container the CJEU annulled part of the Commission's decision for having imposed on the parties to the infringement an obligation to renegotiate or terminate the service contracts concluded between the shippers and the maritime conference, which were not found to be illegal under Article 101 TFEU (as only the provisions of the TAA relating to price-fixing and capacity were found by the Commission to infringe this provision). 30 The Commission had adopted this requirement of re-negotiation or termination as a prophylactic remedy in order to prevent the members of a cartel to continue to apply unlawfully fixed prices simply because these prices were incorporated in long-term contracts. In other words, the requirement to renegotiate or terminate the service contracts was justified by the fact that the effects of the infringements identified in the decision would have continued to exist if the addressees of that decision were able to continue to enjoy the economic advantages secured by ongoing contracts entered into on the basis of the horizontal agreement to fix prices and limit supply found illegal by the Commission. 31 As this section of the decision formed part of the order bringing the infringement to an end, the Commission alleged that there was no need to give specific reasons or to draw it to the attention of the parties concerned in the statement of objections. The Court did not agree with this view, noting that the likelihood of private actions for damages should be a sufficient disincentive for the defendants to continue behaviour that would have maintained or facilitated the effects of the core infringement to Article 101 TFEU, and in any case, the Commission had the obligation to ‗explain its reasoning' as to how the prophylactic measures suggested were ‗obviously necessary' to put the main infringement to an end, something that the Commission's decision had not done. 32 The Court even observed that ‗the statement of objections should in any event have set out, even briefly, but in sufficiently clear terms, the measures which the Commission intended to take in order to bring an end to the infringements and should have given the applicants all the information necessary in order to enable them properly to defend themselves before the Commission adopted a final decision on that point', in view of the rights of defence of the defendants and the requirement that they should be offered a proper opportunity to make known their view. 33 The instrumental use of the principle of effectiveness in order to expand the remedial discretion of the competition authorities is closely interrelated to the question of the legitimacy of competition authorities and the judiciary in competition law cases to adopt far-reaching remedies that may restrict individual rights 34 . Both public and private lawyers recognize the existence of inherent limits to the remedial action of 29 Ibid., para. 178. 30  competition authorities and the judiciary, even if they may take different perspectives in this context.
The private law account of remedial action emphasizes correlativity as an essential feature of private law disputes 35 . Correlativity perceives the parties' relationship as a -bipolar unit in which each party's normative position is intelligible only in the light of the other's‖ 36 . The duty of one party is the mirror image of the other party's right.
The structure of the remedy should thus -reflect the structure of the injustice, retracing and reversing the movement between the parties‖. In private law disputes losses by the claimants are correlative to gains by the infringer. Courts may effectively exercise their adjudicatory function to determine, based on the evidence heard on the structure of the pre-existing relation between the claimant and the defendant, the appropriate relational structure post-infringement, with the understanding that this should be equivalent to that prior to the infringement. The court's or other decision-maker's discretion is thus bound in the context of private law resolution of disputes by the fact that remedies should not go beyond the structure of the pre-existing relation between the correlative entitlements and obligations of the parties to the dispute.
An important difference between the private and the public law accounts of the remedial process is that while the former is based on the idea of correlativity, the latter is by essence polycentric, because of the variety of interests to be considered by the public authority charged with the protection of the general interest, by essence a polycentric concept. In a public law context, because the judge or authority is seeking to implement generally articulated, aspirational norms in highly differentiated contexts, liability norms do not dictate the content of the remedy. Liability norms only provide the goals and boundaries for the remedial decision. In this context, the linkage between remedies and rights (or wrongs) is instrumental, as the liability stage indirectly constraints the targets of the remedial process, whose aim is to provide a solution to a specifically determined (at the liability stage) problem 37 . The normative parameters that have been set at the liability stage in the form of problems that the remedy must address operate at the same time as a guide and as a constraint to the exercise of remedial discretion in a public law context. Some trade-off devices are thus developed in order to mediate between the different interests represented in these polycentric disputes, the principle of proportionality being one of them.
It follows from the above that despite the instrumental use of the principle of effectiveness in order to elaborate a minimum core of -hybrid‖ EU/national remedies, with the aim to ensure the effective enforcement of competition law, legitimacy concerns may limit the remedial action of competition authorities and courts. The following section attempts to address the question of legitimacy in the context of remedial action, by exploring the hypothesis that this is related to the limits of adjudication, as a distinct form of social ordering than, for instance, regulation or contract. 35

III. Legitimate remedies and the limits of adjudication
In exploring the limits of adjudication, with the aim to set clear limits to the legitimacy of judicial action, Fuller referred to the principle of polycentricity, which he defined as -situation(s) of interacting points of influence‖ which -involve many affected parties and a somewhat fluid state of affairs‖ 38 . Intervention in this context may have -complex repercussions‖ 39 . Fuller observes that one may -visualize this kind of situation by thinking of a spider's web. A pull on one strand will distribute tensions after a complicated pattern throughout the web as a whole […] This is a ‗polycentric' situation because it is ‗many centered'each crossing of strands is a distinct center for distributing tensions‖ 40 .
According to Fuller, the adjudication of polycentric disputes is problematic because the complexity of the dispute and the range of those affected, which sometimes it is difficult to foresee, render it quite difficult to organize their participation to the dispute so as to represent their position. Informed only by the litigating parties, the decision-maker is ill-equipped to determine the impact of the decision reached on the different interests affected, with the consequence that the decision reached may negatively affect societal welfare. Because of the adjudicator's limited competence, Fuller considers that in view of the need to appreciate complex repercussions to interests not directly represented in the dispute, significantly polycentric disputes.
What becomes important is thus the -question of knowing when the polycentric elements have become so significant and predominant that the proper limits of adjudication have been reached‖ 41 . Problems that are sufficiently polycentric may be unsuited for adjudication and may be resolved through managerial direction, through negotiation and contract 42 , or left to the forces of the market. Some problems, such as the allocation of economic resources, may indeed -present too strong a polycentric aspect to be suitable for adjudication‖ 43  This does not necessarily mean that contracts and managerial administrative discretion do not have problems of their own when dealing with polycentric disputes 45 . Contracts are generally ill-suited to inequalities of bargaining power, and managerial and administrative discretion may raise important problems of unlimited discretion. Yet, these methods may integrate more fully the under-represented interests to the adjudicatory procedure and take into account evidence ad arguments emanating from others than the initial parties to the dispute.
Fuller's reference to polycentric disputes is highly relevant for our discussion of competition law remedies, the remedial process constituting an instance of adjudication. The enforcement of competition law requires the balancing of different interests, even though recourse to balancing (or the rule of reason as it is called in US antitrust law) is in some circumstances limited by categorical thinking and the establishment of presumptions (per se categories in US antitrust law or object restrictions in EU competition law). In fine, the formation of specific categories, such as specific types of abuse, subject to particular standards or tests (i.e. margin squeeze, refusal to deal), depends, however, on some pre-balancing of the interests usually affected, together with considerations of procedural economy and error cost analysis. Hence, categorization may be considered as a form of dynamic balancing of principles and policies with the aim to promote the effectiveness of competition law enforcement: in this case reduce the administrative costs of enforcement while avoiding the risk of substantive errors. An illustration of this dynamic approach in establishing legal categories constitutes the evolution of the object box in EU competition law, which has recently been expanded to cover certain forms of information exchange 46 , the prohibition of Internet sales in the context of a vertical agreement 47 , or even vertical practices particularly suspicious in view of the structure of the specific market 48 , the latter being considered by some as an excursion outside what has been considered so far as the traditional domain of the object box, which does not require the consideration of information relating to the structure of the relevant market affected.
The move towards a more effects-based approach has accentuated the polycentric dimension of competition law disputes. Final and intermediate consumers active in the specific relevant market were added to the list of the participants affected by the adjudicated transaction, their interest(s) being given the most important weight in the decision-making process (because of the emphasis on consumer harm). and telecommunications, which may have an important impact on the economy overall. The remedies adopted in the context of such disputes have often been characterized as an instance of -regulatory antitrust‖, the perception being that competition law has moved away from its archetype of adjudication towards a different form of social ordering, such as regulation 49 . There seems to be a fine line between competition law remedies and regulatory remedies, which in some cases, EU competition law has seemingly transgressed, thus leading to the accusation of -regulatory antitrust‖, the latter perception being perceived as an oxymoron. For instance, Ibañez Colomo, argues that this transgression may take different forms: First, competition may become regulatory in nature ‗when its application on a proscriptive basis (rather than prescriptive basis) would not be possible given the features of the relevant market'. 50 Second, the ‗expected standards of intervention in a competition law case can be defined as the composite of (i) the gravity of the infringement identified and (ii) the remedies (if any) required to bring an end to it', the relationship between the two being presumed to be a ‗linear one, that is, the intrusiveness of a given remedy increases in direct proportion with the gravity of the infringement'. 51 However, when competition law is -instrumentalised‖, and this expression is left without definition by the author, -the remedies imposed in a given case may exceed what would be necessary to bring an end to the infringement identified by the authority‖ 52 . The correlative relation between remedies and rights/wrongs is thus perceived as the defining boundary between competition law remedies and regulatory remedies. The two -do not follow the same logic‖, as the former are generally concerned -with preserving the existing market structure from being further deteriorated‖, thus leading to remedies that reflect the wrong committed, and not with re-designing the current market structure, by -sharing (or neutralising competitive advantages)‖, which may give rise to remedies that go beyond the wrong committed. 53 In an economically-oriented competition law, this boundary is more easily described than practised. Theories of consumer harm may not only relate to the structure of the supply side but may also be generated by the specific characteristics of the demand side. Behavioural economics may provide insights on how some behavioural biases of consumers may be exploited by incumbents in order to raise prices. If the practices of the incumbents are caught by competition law, the remedy will need to address these behavioural biases in order to be effective. Yet, providing for remedies dealing with existing behavioural biases will not just restore the competitive process, but will generate a very different one than the one prior to the identified ‗competition law' infringement. hence being outside the normal scope of competition law remedies as it aims to shape the conditions of competition in the market?
Yet, the polycentric character of complex competition law disputes is much more limited than that of pure regulatory law disputes. First, competition law remedies 54 relate to the exercise of an adjudicative function, either of a competition authority or of a court, defined as the adversarial presentation of proofs and reasoned arguments to a decision-maker 55 , and should thus be distinguished from remedies adopted in the context of a rule-making activity, as is often the case in regulation. Second, even when regulators exercise an adjudicative function in enforcing regulation, the polycentric nature of the regulatory dispute is more pronounced than in the context of competition law, the decision taking explicitly into account the economic conditions of an entire sector of activities, rather than the competitive conditions of a specific relevant market on which the parties to the dispute are active, by definition a narrower exercise. Third, the interests that are usually considered in a regulatory dispute are in principle more diverse than those taken into account in competition law disputes, the regulators assuming various responsibilities, such as the protection of the environment, universal service, security of supply, and so on, while competition authorities' role is primarily confined to the protection of the competitive process. As a result of the variety of regulatory goals, there is more extensive participation in the decision-making process of actors representing diverse interests, often not directly related to the dispute.
Focusing, for illustration purposes, on merger control, which is the closest a competition law procedure can come to a regulatory law one, the EU Merger Implementing Regulation provides for the participation in the process of ‗third parties', a category which is narrowly defined as including those having a ‗sufficient interest' in the Commission's procedure, such as customers, suppliers, competitors, members of the administration or management organs of the undertakings concerned or recognised workers' representatives of those undertakings. 56 Certainly, the Commission has appointed a Consumer Liaison officer and might also invite the views of other interested third parties including consumer organisations, 57 but these parties do not have a right to be heard in the absence of an explicit invitation by the Commission. In any case, the third parties are expected to comment only on the competition implications of the merger, rather than on broader issues, such as the protection of employment, environment, and so on. 58 This contrasts with the wide participation of various interests in the context of regulatory decision-making, often with the involvement of intermediary, although not elected, bodies representing a supposed public interest, and less frequently, by direct intervention from interested 54 With the notable exception of remedies imposed in the context of market investigation references in the UK, following publics. Consequently, despite the polycentric dimension of most competition law decisions, 59 remedies are precisely confined to address the specific situation under adjudication. Indeed, the boundaries of remedial discretion are delimited by the interplay of the specific goals entrusted to competition authorities, the principle of remedial proportionality and the control of legality for misuse of powers. 60 One may however criticize Fuller's narrow view of adjudication as essentially involving adversarial presentation of evidence and arguments between the parties to the litigation. First -concealed polycentric elements‖ exist in all problems solved by adjudication 61 . Second, dispute resolution does not constitute the only model of adjudication. Commenting on the development of constitutional adjudication, Hiss argued that a new model of adjudication, which he called -structural reform‖ has emerged challenging the narrow conception of the function of adjudication 62 . While the dispute resolution model is a -sociologically impoverished universe‖, -one that does not account for social groups and bureaucratic institutions‖, its world being composed exclusively of individuals, the litigants, the -structural‖ model is characterized by a multiplicity of parties and an array of competing interests and perspectives 63 . Although dispute resolution -implies a unity of functions in party structure‖, i.e. -the plaintiff is simultaneously the victim, the beneficiary of the remedy, and also the spokesperson‖, the structural model is characterized by a fragmentation of the roles, precisely because of the introduction of sociological entities, such as in the context of competition law disputes, the competitors or consumers. For instance, the possibility for the Commission, acting on its own initiative, pursuant to Article 15(3) of Regulation 1/2003, to submit written observations ("amicus curiae" observations) to courts of the Member States where the coherent application of Article 101 or 102 TFEU so requires, provides the opportunity to the national courts to hear evidence and arguments that may not have been proffered by the litigants and which relate to the broader EU interest to ensure a coherent enforcement of EU competition law throughout the Union. The function of remedies is also different in the context of the -structural‖ model. Its main role consists in eliminating threats to the values (effective competition) protected by the law (a prophylactic aim), eventually restructuring the organization that committed the infringement, -a complex and difficult task wholly alien to the dispute resolution model‖ 64 . Consequently, while the remedial process in the dispute resolution model is based on the assumption that the aim of adjudication is to assess the -abnormal‖ event that has caused the dispute and to restore the parties to their rightful position, that is, the position that they would have occupied absent that specific abnormal event (in other words the violation of EU competition law), the -structural‖ model -reflects doubt on whether the status quo is efficient‖, eventually promoting the establishment of a new status quo 65 . 59 Which do not only affect the parties to the dispute but also other competitors, consumers, customers, shareholders, employees. 60 This ground of judicial review refers to cases where an authority uses its powers to take a measure with a purpose other than that for which the powers in question were conferred to it. Behavior 121-128. 63 Ibid., 123. 64 Ibid., 123. 65 Ibid., 124.
One should therefore come into grips with the fact that competition law enforcement constitutes an instance of -structural‖ adjudication, rather than belonging to the -dispute resolution‖ model. This renders the distinction with the administrative/managerial model particularly difficult at times, at a practical level, although there is some value in thinking of the two (adjudication and administrative managerial model) as ideal types forming a continuum with regard to the -appropriate‖ degree of discretion and consequently the legitimacy of the action of the institutions in charge, which is closely related to the participation of the interests affected. Put simply, the more EU competition law moves towards the regulatory/managerial model, and -structural‖ adjudication comes close to that, the more it should integrate the legitimacy-building mechanisms of such model, with an enhanced participation of the entities subject to the remedies as well as of all those whose interests may be affected (i.e. consumers, competitors in related markets, interests vicariously represented by organizations and citizen's groups, i.e. environmental associations). The triadic model of dispute resolution, limited to the parties and the adjudicator, needs to give way in circumstances of significantly polycentric disputes to the more inclusive model of -structural‖ adjudication in order to preserve the legitimacy of competition law enforcement and its continued relevance and appeal to society at large.

IV. Remedial discretion and commitment decisions: exploring the limits of adjudication
The delimitation of the fine boundary between adjudication and the managerial or administrative model is not the only challenge to which mature competition law enforcement systems are confronted in devising effective remedies. In recent years, we have witnessed a competition law enforcement, and the need to adopt procedures that maximize the aims of competition law enforcers, while keeping an eye on the legitimacy of their action.
One may validly ask if commitment decisions fit the adjudicative model or may be considered as closer to the model of negotiation and contract, or that of the regulatory model, which form distinct categories of social ordering if we follow Fuller's perspective.
One could explore the similarities and differences between commitments decisions in antitrust and in merger control, in order to characterize such decisions as fitting in one of the models of decision-making advanced by Fuller. In merger control, remedies take the form of commitments offered by the parties to the merger, either at Phase I or Phase II of the merger procedure, which are eventually accepted by the Commission if they address its ‗serious doubts' over the legality of the merger or the ‗competition concerns' identified. This leads to the publication of a decision under Article 6(2) or 8(2) of the EC Merger Control Regulation, which makes binding the commitments offered by the parties. In the context of the ex post competition law enforcement of Articles 101 and 102 TFEU, Article 9 of Regulation 1/2003 enables the Commission to make commitments offered by parties to its proceedings binding on them, instead of issuing a regular prohibition decision, when those commitments address the concerns expressed in the Commission's preliminary assessment. Such a decision may be adopted for a specified period and ‗shall' conclude that there are no longer ‗grounds for action' by the Commission. Technically, commitment decisions offered under Article 9 of Regulation 1/2003 are not remedies as they do not aim to put the infringement to an end, as it is the case for Article 7 of Regulation 1/2003 and phase II merger control decisions and do not make any finding as to whether there has been an infringement. 69 Their only legal effect is to close the Commission's proceedings. Essentially, it is considered as a measure of -procedural economy‖. 70 In addition, as it was noted by AG Kokott in Alrosa, -unlike Article 7, Article 9 of Regulation 1/2003 is not an instrument for establishing infringements of competition law, but merely gives the Commission the possibility of effectively addressing concerns over competition for the future‖. 71 Contrary to Article 7 infringement decisions, they cannot be used as conclusive evidence of the existence of an infringement of EU competition rules in follow on private actions for damages. 72 Yet, from a functional perspective they can be qualified as ‗remedies', as they aim to redress the situation of the victims of the competition law violation to that prior to the infringement.
As both Article 9 of Regulation 1/2003 commitment decisions and decisions adopted in the context of merger control are formally suggested by the parties to the transaction, they have been distinguished from other competition law remedies and sanctions, which are imposed unilaterally by the Commission and are not the product of a -voluntary‖ agreement between the Commission and the parties to the dispute (coercive remedies). 73 In a similar vein, commentators, and most recently the CJEU, consider that commitment decisions form part of what has been characterized as a -consensual competition law enforcement‖ or a culture of -settlement‖, hence accentuating the opposition between the voluntary nature of commitment decisions and the coercive nature of Article 7 of Regulation 1/2003 decisions imposing injunctions on the parties. 74 The EU courts have relied on the classification of remedies as voluntary or coercive, when dealing with the question of the degree of the remedial discretion competition authorities benefit from in EU antitrust and merger proceedings. Competition authorities are subject to restrictions in the use of voluntary remedies, at least in antitrust proceedings. Recital 13 of Regulation 1/2003 warns that commitment decisions under Article 9 may not be appropriate in cases where the Commission intends to impose a fine. Hardcore cartel cases, generally subject to fines, cannot be closed by a commitment decision. 75 The principle of proportionality may also limit the remedial discretion of competition authorities in both merger and antitrust proceedings, thus reinforcing the legitimacy of the remedial action of competition authorities.
Yet, in Alrosa, the CJEU struck down the judgment of the General Court for having applied to a similar extent the proportionality control in Article 9 and Article 7 decisions. 76 The CJEU noted that ‗the obligation on the Commission to ensure that the principle of proportionality is observed has a different extent and content, depending on whether it is considered in relation to the former or the latter article'. 77 It further explained that: -application of the principle of proportionality by the Commission in the context of Article 9 of Regulation No 1/2003 is confined to verifying that the commitments in question address the concerns it expressed to the undertakings concerned and that they have not offered less onerous commitments that also address those concerns adequately‖. 78 Although both Article 7 and 9 decisions are subject to the principle of proportionality, the application of that principle differs according to which of those provisions is concerned. Hence, according to the CJEU: Following Alrosa, the distinction between voluntary and unilateral remedies leads to a different application of the proportionality principle, hence to a greater variation in the degree of judicial scrutiny of remedies and consequently the remedial discretion of the Commission. In addition to the differential implementation of the proportionality principle, in the context of judicial review, the perceived consensual nature of the commitment procedure may weaken the link between the wrong and the remedy adopted, presumably pushing the authority to use its bargaining power in order to achieve remedies that would not only attempt to reverse the situation to the status quo ante the infringement but would also aim to establish a new, allegedly more competitive, equilibrium, under the pretext that the parties subject to the remedy have consented to it. Indeed, the voluntary nature of commitments may enable them to go further in their scope than what could be the scope of an infringement decision, if the latter option was chosen 80 .
The distinction between infringement decisions under Article 7, allegedly inspired by a -public law‖ paradigm and commitment decisions, relying on a -contract law paradigm‖ does not however take sufficiently into account the important similarities between injunctions and commitments in EU competition law. The reference to the ‗public law paradigm' as a separate pole to the ‗contract law' one seems far-fetched, in view of the importance of ‗administrative contracts' in continental administrative law, but also of the distinction between imperium merum (the power to coerce) and jurisdictio (the power to make legal decisions). 81 Remedies do not form part of the imperium but of the mixtum imperium, the power which a magistrate has for the purposes of administering the civil (not criminal) part of the law, which is incident to 78 Ibid., para. 41. 79  jurisdictio. If remedies were classified within the imperium it would not have been possible, first, for arbitration clauses to be included in merger remedies, arbitration being in this case a forced -contract‖, which is a distinct possibility in EU merger control, 82 and, second, for remedial injunctions to produce extraterritorial effects. 83 More troubling is the opposition sometimes made between the passive role of the parties in Article 7 proceedings and their active role in commitment decisions, in merger control or in the context of Article 9 of Regulation 1/2003. Despite the -coercive‖ appearance of an injunction, often this is the result of a prior (failed) negotiation between the Commission and the parties concerned, the Commission attempting at least to achieve some form of adhesion from the parties that will guarantee the proper execution of the remedy. 84 The psychological pressure that an infringement decision might be adopted by the Commission, in the absence of commitments offered by the parties, largely denies the voluntary and consensual nature of the process and enables the Commission to extract disproportionate remedies.
To this one may add the possibilities offered to the Commission to make strategic use of commitment decisions in order to achieve regulatory objectives, such as the liberalization of the energy sector, the regulation of payment systems, or the opening up of systems competition in the high tech or the car industry sectors, to cite a few examples. As I have shown elsewhere the parties -consent‖ to these commitments, despite the obvious resources-related constraints of the Commission to follow all possible cases with an infringement decision, in view of the collective action problem they are facing, thus enabling the Commission and other competition authorities to leverage their investigative powers into substantial bargaining power, leading to commitments that do may go beyond what would have been achieved, had the Commission chosen instead an Article 7 infringement decision 85 .
The competition authority may also adopt the strategy to invest first in some high profile Article 7 infringement decisions for certain kind of practices, which it can later capitalize with a number of commitment decisions anchored to this quite favourable legal precedent of an infringement decision, pushing the parties to commit to farreaching remedies. A closer look to the commitment decisions adopted by the European Commission in recent years indicates that these relied on contested theories of harm, which may have been risky for the Commission to rely upon in an Article 7 infringement decision: excessive prices 86 , patent ambush 87 , collective dominance 88 , strategic underinvestment 89 , anticompetitive use of court injunctions 90 , to cite a few.
As negotiations occur at the highest levels, between the Competition Commissionaire and the CEOs of these global, in most cases, corporations, commitment decisions may offer some room for politics and the consideration of broader public interests and values than the customary narrow focus on competition concerns, hence highlighting the polycentric nature of most of these disputes. For instance, although issues of privacy were not as such touched upon in the Commission's competition investigation against Google, it is inevitable that the Commission's investigation and the final decision on Google's commitments, which should be taken by the college of commissioners will be influenced by the overall context of Google's disputes with European regulators 91 and, likely, by the European policy to form a digital single market 92 . Broader competition policy concerns, such as market liberalization, were particularly influential in the choice of the instrument of commitment decisions in various cases the Commission investigated in the gaz and electricity markets 93 . Similar concerns over the need for regulation may have played in the financial services sector 94 . Finally, one should also factor in the considerable technical complexity of some markets investigated by the Commission, in particular in the high tech sector. It is virtually impossible to develop sound and effective remedies without an extensive and ongoing cooperation and sharing of information with the business entity which was found to infringe competition law and asked to restructure its organization and/or modify its conduct 95 . This may be best achieved in the context of a commitment procedure, rather than an Article 7 infringement decision, as the atmosphere of open conflict that the latter may generate may be an impediment to cooperation.
The above, and not reasons of procedural economy, constitute the major drivers for the increasing use of commitment procedures in EU competition law.
re-structuring of the organization involved in the violation of competition law, and not just the issuance of a prohibition aimed at some specific act or conduct 96 . For instance, in the recent Visa commitment decision, in addition to the various price caps on the setting of multilaterally agreed interchange fees to which Visa agreed, it also took the commitment to implement further transparency measures re-organizing its relations with the merchants' banks (acquirers) 97 . The process is not also entirely at the hands of the competition authority and the infringer, as one would have expected to happen, in view of the -contractual‖ nature of the procedure, but involves the participation of interested third parties. These may contribute, according to Article 27(4) of Regulation 1/2003, to the -market test‖ of the commitments, offering their view on the commitments offered by the defendant, before these are to be made binding by a decision of the Commission. In order to enhance the transparency of the process to the public, the Commission also publishes the full text of the commitments, as well as a press release setting out the key issues of the case. Interested parties may submit their observations within a fixed time limit which cannot be less than one month. It is also mentioned in the best practices of the Commission, that it may also -actively promote the market test‖ by sending, for instance, the market test document to third parties -which can potentially be concerned by the outcome of the case (e.g. consumer associations)‖ and by -informing in writing the complainant, inviting it to submit comments‖ 98 .
Yet, despite these efforts to enhance participation in order to address the legitimacy concerns raised by the extensive use of the commitment procedure, the Commission enjoys a significant degree of unfettered discretion, in comparison to the one it disposes when adopting an infringement decision. First, the Commission enjoys a wide discretion to make the commitments binding or to reject them, without that choice being framed by self-restraining guidelines, as it is the case in some Member States 99 . Second, there is no specific timeframe for accepting commitment decisions, 96 For a similar conclusion, see Skyscanner Limited v. CMA [2014] CAT 16, where although the CAT cites approvingly the CJEU's view that commitments are offered -voluntarily by the parties‖ (para. 40), it also noted that -(t)his is not an infringement finding and we agree with the CMA that this is more in the nature of a regulatory decision subject only to the normal control of judicial review‖ (para. 160, emphasis added). Although the CAT does not seem to note the distinction between the models of regulation and structural adjudication, the implications of this re-characterization of commitment decisions with regard to participatory rights for third-parties are obvious. 97 Case AT.39398 -VISA MIF (2014). 98 DG Competition, Best practices on the conduct of proceedings concerning Articles 101 and 102 TFEU, available at http://ec.europa.eu/competition/consultations/2010_best_practices/best_practice_articles.pdf , para. 116. 99 In comparison, Article 31D(1) of the UK Competition Act 1998 requires the publication of guidelines as to the circumstances in which it may be appropriate to accept commitments and provides that the Competition Authority -must have regard to the guidance for the time being in force‖. More detailed guidance is provided in Part 4 and Annex A of the OFT's Enforcement Guidance, still valid for the purposes of the Competition and Markets Authority (CMA), which, inter alia, note that the -OFT will not accept binding commitments in circumstances where compliance with and the effectiveness of any binding commitments would be difficult to discern and/or where the OFT considers that not to complete its investigation and make a decision would undermine deterrence‖. The recently published Guidance of the CMA on its investigation procedures also provides that -10.15. (t)he CMA is likely to consider it appropriate to accept commitments only in cases where the competition concerns are readily identifiable, will be fully addressed by the commitments offered, and the proposed commitments can be implemented effectively and, if necessary, within a short period of time‖. The Guidance also provides that -10.17. (t)he CMA is very unlikely to accept commitments in which may also be agreed upon even after the statement of objections was sent to the parties 100 , while some national competition authorities limit explicitly the exercise of that discretion by imposing a specific timeframe 101 . Thirdly, third interested parties and the complainant are not offered adequate opportunities to participate to the procedure and represent their own arguments and evidence on the anticompetitive conduct in question. Certainly, according to Article 27(4) of Regulation 1/2003, interested third parties may submit observations after the Commission has published a summary of the case, as well as the main content of the commitments or of the proposed course of action, in the Official Journal and the full text of the commitments is published on the DG Competition website, within a time limit fixed by the Commission, which may not be less than one month from the date of publication. However, the interested parties do not have access to the full evidence collected by the Commission, in particular its preliminary assessment, which may affect their ability to form a more accurate view over the anti-competitive conduct in question and its impact on their interests 102 . Fourthly, interested third parties have limited access to the courts in order to challenge commitment decisions. Third parties constitute -non-privileged applicants‖ and must satisfy the quite strict criteria of Article 263 TFEU, in particular demonstrate that they are individually concerned by cases involving secret cartels between competitors or a serious abuse of a dominant position‖. The concept of -serious abuse of a dominant position‖ is further explained by the OFT's Enforcement Guidance, where it is mentioned that -(w)hen making its assessment, the OFT will consider a number of factors, including the nature of the product, the structure of the market, the market share(s) of the undertaking(s) involved, entry conditions and the effect on competitors and third parties. The damage caused to consumers whether directly or indirectly will also be an important consideration‖, this assessment being made on a case by case basis, -taking into account all the circumstances of a case‖. Notwithstanding this individual assessment the OFT Enforcement Guidance recognizes that. As a general rule, the OFT will regard predatory pricing as a serious abuse‖. Furthermore, the UK Competition Appeal Tribunal (CAT) has set some broad standards governing the exercise of the discretion of the UK competition authorities to discuss binding commitments, requesting the consideration of the broader interests of deterrence, transparency, but also of those of third parties or consumers, commitments being explicitly excluded -in serious cases of infringement or where deterrence would be undermined‖: Compare with the recent CMA Guidance on its investigative procedures (2014), which provides that -10.18. (a) business under investigation can offer commitments at any time during the course of that investigation, until a decision on infringement is made‖, although it is also made clear that the CMA is unlikely to accept commitments at a very late stage in the investigation, in particular after the CMA has considered representations on the Statement of Objections‖. 102 Furthermore, EU law establishes a distinction between the right of access to the public file conferred to the defendants or undertakings concerned (in the context of Article 9 commitment decisions), third interested parties involved in the proceedings (such as complainants and intervening parties) and third parties that are not involved in the competition authorities' proceedings. The distinction between these three categories is deemed of particular importance. For instance, in Alrosa, the CJEU noted that even if it constituted previously a defendant in proceedings opened against it and De Beers under Article 101 TFEU, Alrosa was only an interested third party as far as the Commission's commitment decision in the investigation against De Beers under Article 102 TFEU is concerned and consequently did not enjoy the same right of access as De Beers to the public enforcement file: Case C-441/07, European Commission v Alrosa Company Ltd [2010] ECR I-5949, paras 88-95. the decision they challenge 103 . It remains to be seen if these restrictive standing requirements will impede interested third parties from challenging in front of the CJEU commitment decisions that affect their own interests 104 . Furthermore, as I have previously noted, in Alrosa the CJEU has distinguished the proportionality ground of review for Article 9 Regulation 1/2003 commitment decisions from that applying to Article 7 infringement decisions, thus transforming it to some form of reinforced control of the rationality (means-end test) of the remedy, even if the CJEU still keeps the proportionality label. The third step of the proportionality test, the assessment of the existence of a least restrictive alternative, which may usually exercise an important constraining effect to the discretion of the Commission, finds itself significantly emptied from its content. The judicial scrutiny cannot constitute an effective check and balance mechanism of the remedial action of the Commission in the context of commitment decisions. Finally, the Commission is not formally obliged to take into account the views presented by the third-parties, but simply to acknowledge them, without explaining the weight it provided, or not, to them in its final decision to accept the commitments -offered‖ by the parties 105 .
The absence of extensive participatory rights for interested third parties (including the complainant) in the context of commitment decisions, despite the polycentric nature of the disputes frequently addressed by them, constitutes one of the major weaknesses of that procedure. In view of the increasing role commitments play in the conduct of competition law enforcement by the Commission, expanding the participation of interested parties (including the complainant) becomes an important site for procedural reform. As it has been recognized in the recently adopted European Competition Network (ECN) Recommendation on Commitment procedures, -the views of other players in the market on the proposed commitments can play an important part in assessing their adequacy to meet the competition concerns and to allow such third parties the opportunity to submit their observations‖ 106 . The Courts may also play an important role in reminding competition authorities of the need to reinforce the position of third parties (and the complainant) in the commitment decisions procedure. For instance, in some of its judgments the UK Competition Appeal Tribunal (CAT) has emphasized the need to provide the complainant a fair and -structured‖ opportunity to comment until informal commitments (-undertakings‖ in UK competition law jargon) had been accepted, the CAT noting that although the commitment procedure does not envisage -an adversarial system‖, -this does not 103  preclude the need to afford the complainant an opportunity to defend its interests during the administrative proceedings‖ and that he is -given an opportunity to be heard before the OFT (Office of Fair Trading) closes its file‖ and/or accepts the undertakings proposed 107 .
The recent judgment of the UK Competition Appeal Tribunal in Skyscanner v. CMA may provide some food for thought 108 . The CAT was seized on appeal to a decision adopted by the OFT, now replaced by the Competition and Markets Authority (CMA), accepting the commitments of some online travel agents and hotel groups by Skyscanner Limited, a third-party operating a price comparison website, also called -meta-search‖ site, assisting customers to compare pricing between various OTAs and hotels, with whom Skyscanner contracts for the inclusion of their offerings in its meta-search results. The main competition concern the commitments accepted by the OFT attempted to address concerned the restriction imposed in arrangements between the specific OTAs and the hotel groups to restrict each OTA's ability to discount the rate at which room-only hotel accommodation bookings were offered to customers (the so called -price agreements‖). The OFT also expressed some concerns with regard to some Most Favoured Nation clauses ensuring that the retail rates for hotel room bookings provided by hotels to OTAs were no less favourable than the lowest retail rate displayed by other distribution outlets (the so called -rate parity obligations), which the OFT found were capable of reinforcing the restriction of competition resulting from the discounting restrictions, although it did not consider if they could on their own be considered as a restriction of competition. The OFT accepted commitments by the parties, who agreed to remove the complete prohibition on discounting room-only rates and to replace these by limited discounting to -closed groups‖ of consumers, basically a membership group which consumers actively choose to join, up to the level of the commission or margin of OTAs. Although, according to the commitments, the OTAs could not publicise information about the specific level or extent of discounts outside the -closed group‖, they could publicise information regarding the availability of discounts on their own websites or to pricecomparison and meta-search websites. Skyscanner alleged that the OFT failed to take into account properly or at all its representations during the consultation (-market-testing‖) process on the impact these commitment had on the meta-search sector on which Skyscanner was active. Skyscanner expressed the concern that the commitments only focused on intra-brand competition, by allowing OTAs to offer discounts, but that allowing hotels to prevent OTAs from publicizing specific information about discounts to consumers outside the OTAs closed group, and hence to meta-search websites, had a negative effect on inter-brand competition (and pricetransparency for consumers), as consumers would be unable to use meta-search sites in order to compare the actual room prices and discounts offered by different hotels, the only information these meta-search websites would be able to provide being that discounts are available, but not the exact level of discounts. Of particular interest for our discussion is the allegation that the OFT violated the statutory duty of the OFT to consult and to consider representations made in response to its consultation notice. Despite the fact that the OFT had acknowledged the concerns expressed by Skyscanner and had even met with Skyscanner representatives to discuss these concerns, and the fact that Skyscanner was a -third party respondent to the consultation‖ and not a complainant in the OFT's original investigation, the CAT found that the consultation process was -defective‖ as it had -failed conscientiously to address the objections raised‖ by Skyscanner and two other respondents to the consultation, which concerned the possible effect of the proposed commitments on price transparency and on meta-search websites 109 . The CAT noted that the fact that Skyscanner was a third-party, and not a complainant, should not affect -the significance the OFT would attach to a material point raised by a respondent‖ 110 , and that the demands made by the OFT to Skyscanner for further evidence as to the impact of the proposed commitments on the meta-search website sector or price transparency -was not acceptable‖ 111 . The CAT noted that the fact that the OFT was pursuing its investigation on the basis that it had identified a restriction -by object‖ may have deprived it -of the ability properly to appreciate the significance of the role of operators such as Skyscanner‖, but this did not excuse the OFT from not investigating -a plausible point further‖, before taking a decision, and, in any case, the OFT should not have insisted on more evidence or supporting material from Skyscanner itself on this point 112 . This finding was accompanied by some derogatory remarks of the CAT, raising the possibility that -the OFT found the Skyscanner objection inconvenient because it threated to upset a carefully constructed edifice that the OFT believed‖ 113 . Consequently, the CAT uphold Skyscanner's appeal on this ground 114 .
The Skyscanner judgment and the emphasis put by the CAT on the consultation process and the need to actively engage with the representations made by third-parties illustrate the polycentric nature of commitment decisions. This specific nature calls for the consideration of the view (and interests) of all those affected by the decision, beyond the parties to the dispute. One cannot therefore apply the same principles as in the adjudicatory model. While conceptually distinct from the model of administrative management/regulation, the -structural model‖ of adjudication offers a more appropriate description of the hybrid nature of commitment decisions than the model of contractual governance. This has implications on the role and conceptualization of the consultation process, or of the intervention of the judiciary, among other issues, in order to guarantee the legitimacy of the decision-making process.  113 Ibid, para. 97. 114 The CAT also noted that -if a consultation response raises an important and obvious point of principle, it is for the authority to examine it further. This is particularly so where the authority has not carried out an analysis of the economic effects of the practices which it proposes to address with its commitments decision and where that decision itself may generate its own economic effects within the market‖, ibid. para. 90.

V. Conclusion
Left unbound, the principle of effectiveness may offer the opportunity to competition authorities to expand their remedial discretion and to re-design market processes and outcomes in accordance with the dominant interpretation of their statutory objectives. This may appear as a sound sacrifice of legal form in favour of -future-oriented‖ remedies that -lay down rules about how markets should behave in the future‖ 115 . The point made in this paper is that, whatever one thinks of the appropriateness of an expansive view of remedial discretion, which is not, in our view, supported by the restrictive interpretation of the principle of effectiveness in EU law, remedial discretion is naturally limited by the specific function exercised by the remedial process chosen or, more contentiously, imposed by the nature of the dispute. Drawing on Fuller's account of the existence of various forms of social ordering, each of them emerging in specific circumstances/context and having its own principles and limitations, I offered some reflections on the possible limits that the essence of each ideal type of social ordering sets to the expansive interpretative potential of the principle of remedial effectiveness. For instance, the limits of adjudication are of different sort than those of contracts/negotiation or managerial/regulatory discretion. For each of these ideal types, legitimacy concerns operate differently and generate dissimilar demands on process or substantive rules bounding discretion. The polycentric nature of competition law disputes calls for flexibility in the choice of the adequate form of social ordering aiming to achieve the objectives set by the legislator. Its specificity also breaks with the classic view of the adjudication model and hints to the prevalence, in a significant number of cases with a pronounced polycentric element, of what has been called the -structural adjudication‖ model, still distinct from the model of regulatory governance. I discussed the nature of commitment decisions as an illustration of the difficulties of classification, without a proper consideration of the functions and respective limits of each form of social ordering. Commitment decisions have erroneously been characterized as closer to forms of contractual governance or, taking a starkly different perspective, to regulation. In reality these characterization disputes offer little and may be as confusing and unpurposeful as discussions over the gender of angels, should one not develop an overall theory on the meaning of each category and the practical implications of such categorization. This study aimed to sketch such theory by insisting on the need to keep an eye on the specific limitations to remedial discretion, of procedural or substantive nature, that have emerged in each form of social ordering with the aim to guarantee the legitimacy of decision-making. It is not impossible for a competition authority to make the choice of one or the other form of social ordering, should it enjoy the power to perform such choice, if this is advancing its purpose. Yet, the authority should also adopt the mechanisms that were put in place in order to respond to the calls for the legitimacy of decision-making in the context of the specific type of social ordering. New tools, such as commitment decisions, which do not fit with existing categories, may require the development of new mechanisms or a different conceptualization of existing mechanisms in order to ensure the legitimacy of decision-making. This creative process is at presently ongoing in EU competition law.