52008SC0405

[pic] | COMMISSION OF THE EUROPEAN COMMUNITIES |

Brussels, 2.4.2008

SEC(2008) 405

COMMISSION STAFF WORKING DOCUMENT Accompanying document to the WHITE PAPER on Damages actions for breach of the EC antitrust rules

IMPACT ASSESSMENT {COM(2008) 165 final} {SEC(2008) 404} {SEC(2008) 406}

TABLE OF CONTENTS

1. Procedural issues and consultation of interested parties 5

1.1. Background, Commission departments involved and chronology 5

1.2. Consultation and expertise sought 6

1.2.1. Review of the status quo for actions for damages in Europe 6

1.2.2. Identification of various possible options to improve compensation of victims of infringements of competition law 6

1.2.3. Opinions of institutional stakeholders 6

1.2.4. Other consultations 7

1.2.5. Study analysing potential impact 8

1.3. Inter-service consultations 8

1.3.1. Informal consultations within the European Commission 8

1.3.2. Formal Inter-Service Consultation 9

1.4. The Impact Assessment Board 9

1.5. Legal context of the proposed initiative 9

2. Problem to be addressed − the “WHY” 11

2.1. The current legal framework for competition damages actions is ineffective: major difficulties for victims to obtain compensation 11

2.2. The scope and scale of the problem 13

2.3. Causes and drivers 16

2.4. Need for EU action in view of the likely evolution of the problem 20

3. Objectives − the “WHAT” 22

3.1. General objective 22

3.2. Specific objectives 23

4. Policy options − the “HOW” 26

4.1. Identifying the policy options to be assessed 26

4.1.1. Method for identifying and grouping the individual measures into bundles to form Policy Options 26

4.1.2. Individual measures not included in the Policy Options analysed 27

4.1.3. Measures regarding interaction with public enforcement 28

4.1.4. Provision of non-binding assistance for quantification of damages 28

4.2. Description of the Policy Options assessed 28

4.2.1. Policy Option 1: legislative measures maximising facilitation of claims and incentives for victims 29

4.2.2. Policy Option 2: substantial facilitation and partial incentives for damages claims by means of legislative measures 30

4.2.3. Policy Option 3: moderate facilitation by means of more limited legislative measures 31

4.2.4. Policy Option 4: moderate facilitation by means of non-regulatory measures 32

4.2.5. Policy Option 5: zero action at EU level (baseline option) 32

5. Analysis of impact 34

5.1. Assessment criteria 34

5.1.1. Benefits 34

a. Ensuring full compensation of the entire harm suffered 34

b. Increased awareness, enforcement, deterrence and legal certainty 34

c. Access to justice 35

d. Efficient use of the judicial system 36

e. A more level playing field 36

5.1.2. Costs 36

a. Litigation costs 36

b. Administrative burden 37

c. Error costs 37

d. Implementation costs 37

5.1.3. Other impacts 37

5.2. Identifying and assessing the impact of each Policy Option 38

5.2.1. Policy Option 1 38

5.2.2. Policy Option 2 41

5.2.3. Policy Option 3 43

5.2.4. Policy Option 4 46

5.2.5. Policy Option 5 48

5.3. Likely costs for public authorities and businesses under the Policy Options 50

5.4. Other impacts, in particular likely implications for the EU budget 51

6. Comparing the options 52

6.1. The Preferred Option 52

6.1.1. Interim results of the impact analysis 52

6.1.2. Identification of the Preferred Option 55

6.1.3. Impact of the Preferred Option 58

6.2. Summary comparison of Policy Options 61

6.3. Proportionality and EU added value of the Preferred Option 62

6.3.1. Proportionality 62

6.3.2. European added value 62

7. Monitoring and evaluation 65

Procedural issues and consultation of interested parties

Background, Commission departments involved and chronology

1. The key issue addressed in the White Paper is the current lack of an effective legal framework for private actions seeking compensation for the damage caused to citizens and businesses as a result of infringements of EC competition law. While public enforcement by competition authorities punishes breaches of competition law, compensation of victims can only be obtained via national courts, in accordance with national procedural rules.

2. The Directorate-General for Competition is the lead service on the White Paper. The other departments involved are: DG Enterprise, DG Justice, Freedom and Security, DG Internal Market, DG Health and Consumer Affairs, DG Economic and Financial Affairs, the Legal Service, the Secretariat-General and the Bureau of European Policy Advisers. The White Paper on damages actions for breach of the EC antitrust rules is listed in the Commission Legislative and Work Programme for 2007.[1]

3. A study on the conditions under which private parties can bring actions for damages before the courts of the Member States for breach of the EC antitrust rules was undertaken by the law firm Ashurst, in response to a tender awarded by the Commission. This “Comparative Study”,[2] submitted on 31 August 2004, found that levels of private enforcement by means of damages claims in Europe are currently very low and that victims face substantial obstacles when trying to obtain compensation in court (see section 2).

4. On 19 December 2005 the Commission adopted a Green Paper and a Commission Staff Working Document on damages actions for breach of the EC antitrust rules. The purpose of the Green Paper was to consult stakeholders and stimulate debate. It identified obstacles to effective antitrust damages actions and set out different options as a basis for discussion. The Green Paper was met with broad interest.

5. In its 2007 Legislative and Work Programme, the Commission endorsed preparation of a White Paper on antitrust damages actions, which would suggest possible measures to follow up the 2005 Green Paper. To assist with preparation of the White Paper and the accompanying impact assessment, on 23 December 2006 the Commission published a call for tenders for provision of an impact study on the White Paper (the “Impact Study”).

6. Adoption of the White Paper is to be followed by a period of public consultation of all stakeholders.

Consultation and expertise sought

Review of the stat us quo for actions for damages in Europe

7. The Comparative Study identified and analysed the obstacles to successful action for damages in the Member States of the European Union. It stated that levels of private enforcement by means of damages claims in Europe are very low. The study found not only “total underdevelopment” of actions for damages for breach of EC competition law, but also “astonishing diversity” in the approaches taken by the Member States.[3] Following submission of the Comparative Study, the Commission met a number of high-level academics and legal experts to discuss issues arising from it.

Identification of various possible options to improve compensation of victims of infringements of competition law

8. The Green Paper identified various major obstacles to effective antitrust damages actions and set out different options on 15 issues, as a basis for discussion. The Green Paper was met with broad interest by stakeholders: it was discussed at a number of conferences and gave rise to nearly 150 formal submissions. This is in addition to numerous comments and material presented at conferences and in various publications.

9. Practically all the responses accepted the complementary role of private actions in overall enforcement of the EC competition rules. Most respondents agreed that victims of infringements of competition law are entitled to damages and that national procedural rules should be such that this right can be exercised effectively. Although the objective of the Green Paper therefore met with a broad consensus, respondents have diverging views about the nature and extent of obstacles to actions for damages, and the appropriate ways to remedy shortcomings. In particular, consumer associations (and law firms known as “plaintiff lawyers”) are generally in favour of far-reaching measures to facilitate and create incentives for private damages actions. Business associations and their advisers often argue more in favour of a moderate approach, some recommending taking no action at all. Academics and representatives from Member States agree on the objective and tend to favour options from the middle of the spectrum.

10. The White Paper is therefore intended as a basis for further discussion on these issues, based on specific suggestions and a detailed impact assessment.

Opinions of institutional stakeholders

11. The European Parliament contributed to the debate, in April 2007, with a Report and a Resolution on the Commission’s Green Paper. The European Parliament called on the Commission “to prepare a White Paper with detailed proposals to facilitate the bringing of private actions claiming damages which addresses, in a comprehensive manner, the issues raised in [its] resolution and gives consideration, where appropriate, to an adequate legal framework”. The issues raised in the Resolution are, in particular, the complementary role of actions for damages (and the issue of removing the joint civil liability of the successful leniency applicant), the asymmetry of information and of resources between the parties, the disclosure of evidence inter partes , the binding nature of NCA decisions, the need for collective actions, definition and quantification of damages, promotion of settlements, the passing-on defence, costs and limitation periods.[4].

12. On 20 October 2006 the European Economic and Social Committee (EESC) adopted an Opinion on the Commission’s Green Paper.[5] The EESC’s Opinion is positive about the Commission’s efforts to facilitate antitrust damages actions and considers that the Green Paper has opened up a broad and welcome debate on the need to make it easier for those injured by anti-competitive practices to recover damages. The Opinion supports measures at Community level to facilitate antitrust damages actions.

13. On two occasions, first on 11 and 12 September, then again on 9 November 2007, DG Competition met a group of experts from the Member States in order to hear their views on a discussion paper drawn up by DG Competition in preparation for the White Paper. Experts represented the Ministries of Justice, the Ministries of Economic Affairs and the national competition authorities of the Member States. Representatives from the EFTA Surveillance Authority, the Norwegian competition authority, the Icelandic competition authority and the Office of Economic Affairs of Liechtenstein were also invited to the meetings.

14. On 6 November 2007 DG Competition also held an informal meeting with a delegation of judges of national supreme courts, courts of appeal, courts of first instance and specialist competition tribunals from 12 Member States[6] to discuss specific issues related to antitrust damages actions, based on the discussion paper prepared by Commission staff.

Other consultations

15. Consultation on the issues raised in the Green Paper extended far beyond the formal period of public consultation. Since adoption of the Green Paper, Commission staff have repeatedly met a wide range of stakeholders and experts, in particular consumer associations, business representatives, lawyers and academics.

16. Commission staff have participated in a large number of events (conferences, expert panels, etc.) to discuss more effective actions for antitrust damages in the EU and their implications. Moreover, the Commission itself co-organised two major conferences which were attended by a wide range of experts and stakeholders and were a valuable source of feedback – one on “Private enforcement in EC competition law” together with the ERA Academy of European Law (9 March 2006 in Brussels), the other on “Cartel enforcement and antitrust damages actions in Europe” together with the International Bar Association (7-9 March 2007 in Brussels).

Study analysing potential impact

17. An impact study on the White Paper on EC antitrust damages actions (the “Impact Study”) was undertaken by the Centre for European Policies Studies , in association with Erasmus University Rotterdam and Libera Università Internazionale degli Studi Sociali Guido Carli , in response to a tender awarded by the Commission.[7]

18. With this tender, the European Commission was seeking the expertise necessary to produce a study on the economic and social impact of a higher level of EC antitrust damages actions in Europe, especially in comparison with the current recognised low level. The Impact Study was officially submitted to the Commission on 21 December 2007.

19. The Impact Study provides the Commission with a thorough analysis to assist it in assessing the impact of a higher level of EC antitrust damages actions in Europe. In particular, it contains an extensive analysis of the existing research regarding private actions for damages, a welfare analysis of a more effective system of private antitrust damages actions in Europe and a cost-benefits/cost-effectiveness analysis of specific measures and comprehensive options that could enhance private antitrust damages actions in Europe. The Impact Study follows the method in the Commission’s Guidelines on Impact Assessment.[8]

20. In terms of final result, the Impact Study found that a range of measures facilitating antitrust damages actions would have a significant positive impact in the EU, namely ensure adequate compensation for victims, enhance deterrence of unlawful conduct and have positive overall economic effects. The Impact Study also identified the measures most likely to minimise any negative effects, such as higher litigation costs and heavier administrative burdens.

21. The Impact Study provided key input for the Commission in preparation for this “Impact Assessment Report” (or this “report”). However, the Commission has conducted its own analysis of the Impact Study and of the wealth of material gathered during the formal and informal consultations. Therefore, the conclusions of this report may, in some instances, diverge from the findings of the Impact Study, when there are sound legal or policy reasons to do so.

Inter-service consultations

Informal consultations within the European Commission

22. Since the White Paper touches upon a variety of issues, some of which are related to the work of other Commission departments, the Directorate-General for Competition ensured close cooperation with all the Commission departments concerned at different levels and at every stage of the preparatory process. This exchange of views before the official Inter-Service Consultation allowed progress in many areas and alignment of positions on more sensitive topics.

23. The sections of the White Paper which relate to issues touching upon the areas of activity of other Directorates-General have been drafted in cooperation with the DGs concerned. Cooperation was particularly close with the Directorate-General for Health and Consumer Affairs, which, together with the Directorate-General for Enterprise, was, inter alia, represented on the Commission evaluation board in the procedure to choose the contractor to carry out the Impact Study.

24. Furthermore, an inter-service group was set up for this Impact Assessment Report and met on various occasions – on 24 November 2006, 13 July 2007 and 10 December 2007. The members of this group were also given an opportunity to comment on the ongoing work of the contractor preparing the Impact Study.

25. This Impact Assessment analyses a wide range of policy options, some of which may, in different policy areas, also be or come to be assessed by other Directorates-General of the Commission. Based on all these assessments, the various initiatives by the Commission will continue to be closely coordinated. Particularly with respect to collective redress, it will, in due course, be appropriate to assess whether and to what extent a horizontal instrument can effectively tackle the issues identified, or whether specific competition law measures (as analysed in this report) are better suited to rendering the victims’ rights to compensation effective.

Formal Inter-Service Consultation

26. A formal Inter-Service Consultation was launched on 9 February 2008. The Directorate-General for Competition took due account of the various comments received during this consultation.

The Impact Assessment Board

27. A draft of this Impact Assessment was submitted to the Impact Assessment Board on 21 December 2007. In its favourable opinion dated 25 January 2008, the Board advised the Directorate-General for Competition to elaborate on certain problem definition issues, to make the identification of policy options more transparent, to further explain the relationship between public and private enforcement and to expand the analysis of certain forms of impact.

28. The Directorate-General for Competition has revised this report in line with the Board’s comments.

Legal context of the proposed initiative

29. In 2001 and 2006 the European Court of Justice recalled that, as a matter of Community law, the possibility of claiming compensation must be open to any individual who suffers harm as a result of an infringement of Community antitrust rules. As the Court emphasised, Community law requires effective remedies which create a realistic opportunity to exercise this right to damages.

30. The main objective of the White Paper is to ensure effective implementation of the finding of the European Court of Justice that the full effectiveness of the Treaty would be put at risk if there was no realistic chance for every individual to obtain damages for loss caused by conduct liable to restrict or distort competition.

Problem to be addressed − the “WHY”

31. The key issue addressed in the White Paper is the current lack of an effective legal framework for private actions seeking compensation for the damage caused to citizens and businesses as a result of infringements of EC competition law. Deficits regarding reparation of harm resulting from an infringement of directly applicable Community rules also exist in other areas of law. In the field of competition, however, the size of the uncompensated harm and the problems encountered are particularly big. This is mainly due to a number of particular characteristics of actions for damages for competition infringements. These include the very complex factual and economic analysis required, the frequent inaccessibility and concealment of crucial evidence in the hands of defendants and the often unfavourable risk/reward balance for claimants.

32. Companies which infringe the EC competition rules usually gain an illegal advantage on the market at the expense of other players and, in particular, consumers. The damage caused by anti-competitive practices, such as price-fixing or market-sharing agreements, can be immense. For example, in France, courts recently confirmed the findings of the Competition Council that a market-sharing cartel of mobile phone operators had led to damages for consumers that were estimated to be in the range of €295 to €590 million during the period from 2000 to 2002.[9] In another recent case, the UK competition authority estimated that collusion between large supermarkets and dairy processors to increase the prices of dairy products had cost consumers around £270 million (approximately €375 million) during a two-year period.[10] A cartel of brewers on the Dutch market is said to have inflated beer prices for consumers by €400 million from 1996 to 1999.[11] Abuse of its dominant position by a broadband provider and wholesaler in Spain[12] prompted a consumers association to claim €458 million in damages before a court as compensation for five years of illegal overcharging.[13] Infringements which concern more than one Member State and longer periods than these examples are bound to have a far greater negative impact. As a matter of fact, EU-wide infringements are becoming more and more frequent.

The current legal framework for competition damages actions is ineffective: major difficulties for victims to obtain compensation

33. In 2001 and 2006, the European Court of Justice recalled that, as a matter of Community law, the possibility of claiming compensation must be open to any individual who suffers harm as a result of an infringement of Community antitrust rules.[14] Community law requires an effective legal framework which creates a realistic opportunity to exercise this right to damages. This framework can consist of Community rules on the matter or, in the absence thereof, the legislation in the Member States on civil law and procedure, or a combination of the two. For the time being, no specific Community rules (other than those derived from the principles of effectiveness and equivalence) exist regarding claims for competition damages before civil courts, so the national rules on civil liability and civil procedure apply.

34. Under these very diverse national rules, in practice citizens and companies encounter difficulties in obtaining compensation for antitrust damages before the civil courts. The traditional legal mechanisms are not working effectively in the specific context of antitrust damages claims. This general problem identified by the Comparative Study, the Commission’s Green Paper and the annexed Staff Working Paper was largely confirmed during the public consultation. By contrast, many respondents who themselves face an increased likelihood of becoming (or representing) a defendant in damages actions considered the current legal and procedural framework appropriate.[15] The Impact Study on measures facilitating antitrust damages actions in Europe concludes that under the current systems it is very difficult to exercise the right to damages and that very few victims receive compensation.[16] The European Parliament and the European Economic and Social Committee also concur with the findings in the Green Paper and have called upon the Commission to address the problem of the ineffective legal framework for antitrust damages actions in the interest of consumers and businesses.[17]

35. The Comparative Study actually found “total underdevelopment” of such damages actions, an “astonishing diversity” in the approaches taken in the Member States and a considerable degree of legal uncertainty across the EU.[18] The study identified only a very limited number of successful damages awards for breach of EC antitrust rules since 1962.[19] The Green Paper and the annexed Staff Working Paper describe the main difficulties and obstacles in the legal systems of the Member States encountered by victims of infringements of competition law when trying to exercise their entitlement to damages;[20] section 2.3 recalls the main causes and drivers. Since the Comparative Study and the Commission’s Green Paper, an increase has been reported, in some Member States, in attempts by victims to claim damages before civil courts. However, the problems identified in the Comparative Study remain essentially unchanged, as in most Member States and across the EU the number of cases where victims have been successful in their action for antitrust damages[21] is still very limited (see section 2.4 for further details).

36. No statistics are available on the number of (usually confidential) settlements between victims and infringers or successful cases of alternative dispute resolution such as mediation. However, deficits are bound to exist in this respect too, as long as there are significant obstacles to judicial enforcement of the right to damages, given that, to a large extent, effective and equitable settlements and mediation presuppose a realistic possibility of alternatively enforcing the damages claim before civil courts.

37. As a consequence of the current situation, a large number of victims of infringements of competition law remain uncompensated for the harm suffered and see their right to damages under the EC Treaty frustrated. There is therefore a clear deficit in terms of corrective justice .

38. In addition, the lack of an effectively functioning legal framework for antitrust damages actions also precludes other beneficial effects of private enforcement of Treaty rights, namely the deterrence of future infringements inherent in effective compensation mechanisms.[22] If infringers of competition law have no fear of having to pay damages to their victims, they have fewer incentives to refrain from engaging in anti-competitive conduct. If victims have access to effective legal mechanisms to bring infringements of competition law before civil courts, there is a greater likelihood that a larger number of illegal restrictions of competition will be detected and addressed in full, as victims could have particular knowledge about the anti-competitive practices and may decide to take the initiative if the competition authorities do not pursue the case (e.g. for reasons of priority-setting because of limited resources). The Court of Justice emphasised this complementary function of private enforcement of directly applicable Treaty rights as early as 1963[23] and, in later years, made it clear that without a working mechanism for victims to obtain redress, the effectiveness of Articles 81 and 82 of the EC Treaty is put at risk.[24]

39. Articles 81 and 82 are a matter of public policy and lie at the heart of the functioning of the internal market.[25] Shortcomings in effective enforcement of Articles 81 and 82 hinder achievement of the results of competition, namely better allocation of resources, greater economic efficiency, increased innovation and lower prices. In this way, such shortcomings also have a direct impact on the functioning of the internal market, which relies on a system of undistorted competition.

The scope and scale of the problem

40. The problem outlined above is not limited to any particular Member State. It is observed in every Member State, although to differing degrees as the applicable national rules differ significantly (see section 2.4). Infringements of competition law, be they hardcore cartels, other infringements of Article 81 or abuses under Article 82, occur in almost every sector of the economy.[26] The problem outlined above concerns both actions for damages brought following a decision by a public authority and actions brought on a stand-alone basis. Moreover, the problem concerns all groups of citizens and businesses. Especially affected, however, are large groups of consumers and small and medium-sized enterprises (SMEs): often at the end of the distribution chain, they face particular difficulties in identifying and proving the harm they suffered (quantum and causal link) and perceive the uncertainties, risks and costs of an action as disproportionate to potential benefits.[27]

41. It is impossible to quantify precisely the total cost of the current ineffectiveness of the legal framework for antitrust damages actions to the direct and indirect victims and to society as a whole. There are no reliable empirical data, neither on the number of infringements of competition law occurring in Europe every year nor on the scale of the harm they cause. Especially for the gravest and most harmful violations, infringers usually go a long way to conceal their practices and their impact.

42. There is, however, general agreement that the total harm done to society by infringements of competition law is immense. Looking alone at hardcore cartels with effects across the EU ,[28] the Impact Study estimates that the annual direct cost to consumers and other victims in the EU ranges from approximately €13 billion (on the most conservative assumptions) to over €37 billion (on the least conservative[29]). This estimate comprises both the harm resulting from consumers and other victims having to pay a higher price due to the illegal conduct of the cartelists (the “overcharge”) and also the economic benefits forgone by consumers and other victims who do not purchase, or purchase a smaller quantity, due to the unlawfully inflated price (the “deadweight loss”[30]). It therefore covers the direct costs of cartels to consumers and other victims. It takes no account of more indirect macro-economic effects, such as the absence of greater allocative, productive and dynamic efficiency, which could have contributed to growth and employment, which are extremely difficult even to estimate.

43. If one adds to the figures on EU-wide cartels the annual cost to consumers and other victims of domestic hardcore cartels, the total annual cost for hardcore cartels in the EU can be estimated to range from approximately €25 billion (on the most conservative assumptions) to approximately €69 billion (on the least conservative).[31] Expressed as a proportion of the EU’s gross domestic product, the negative consumer welfare impact of all these hardcore cartels is estimated as ranging from 0.23% to 0.62% of the EU’s GDP in 2007,[32] which does not include the harm caused by abusive practices and infringements of Article 81 other than hardcore cartels. To illustrate the harm created by cartels and thus the potential benefits of enhanced private enforcement differently: if more effective compensation mechanisms were to lead to a reduction of hardcore cartels by, for example, 5%, the negative consumer welfare impact would be reduced by €1.25 to €3.45 billion.

44. Even in the most effective system of private enforcement, not all the harm to consumers and other victims reflected in the above estimates will be compensated: this is because, inter alia , a considerable number of antitrust infringements will remain undetected. For hardcore cartels, the detection rate is generally assumed to be no more than somewhere between 10% and 20%.[33] For other infringements, the detection rate is higher, but the “conviction” rate (i.e. the rate of successful damages actions) is likely to be much lower, since claimants often find it very difficult to produce proof that the contested conduct produced actual anti-competitive effects.[34] It also has to be assumed that some victims do not come forward to claim compensation, for instance because they prefer not to disrupt an ongoing business relationship with the infringer. Moreover, in some instances, victims will find it rather difficult to convince courts of a sufficiently close causal link between any particular damage and the infringement.[35]

45. Any realistic estimate of how much compensation victims could expect under a more effective legal framework for antitrust damages actions is therefore necessarily lower than the total harm to consumers and other victims from detected and undetected infringements estimated (for cartels) above. The Impact Study assesses the potential benefit of a more effective compensation system in the EU by comparing the current ineffective legal framework in Europe with a legal system where private enforcement of competition rules by means of damages actions is very effective, i.e. where victims of antitrust infringements no longer encounter the same obstacles to claiming compensation in court for the damage suffered.[36] On this basis, the Impact Study estimates that the total amount of compensation (single damages plus pre-judgment interest[37]) that victims of antitrust infringements are currently forgoing ranges from approximately €5.7 billion (on the most conservative assumptions) to €23.3 billion (on the least conservative) each year across the EU.[38] These estimates relate to all types of infringements of Articles 81 and 82. They provide an approximate idea of the amount of compensation that victims are currently forgoing. However, they should not be taken as a precise calculation of the magnitude of future antitrust damage awards, given that they have to be based on a range of, sometimes big, assumptions. They can only be taken as just an indication of the potential for more effective private damages actions.

46. More effective antitrust damages actions implies more cases. An increase in the number of court actions will result in additional costs, in particular for businesses, consumers and the judicial system. It is not possible to quantify with a sufficiently satisfactory degree of precision the costs to be expected as a result of this increase. However, sections 5 and 6 of this report assess alternative policy options in terms of the likely litigation costs, administrative burden and error costs, for both private and public entities.[39] It is even more difficult actually to measure, taking account of all potential benefits and costs, the net gains (or reductions) in terms of possible efficiencies (greater deterrence, lower prices, better quality and resource allocation and possible macro-economic effects) that could result from an effective legal system for private antitrust damages actions. Nevertheless, one finding of the Impact Study is that, in any scenario, the costs resulting from an increased level of damages cases would not offset the potential benefits in terms of compensation of victims as a result of the higher level of actions for damages.[40] Sections 5 and 6 set out a more detailed comparison of the positive and negative impacts likely to be triggered by alternative policy options.

Causes and drivers

47. During the Commission’s various formal and informal consultations of stakeholders, it became clear that the ineffectiveness of the legal framework for antitrust damages actions is due to a set of different causes and drivers: it is partly the result of the high degree of legal uncertainty that potential claimants, but also defendants, face. It is also due to a range of legal and procedural hurdles in the traditional rules of the Member States. These factors, combined with the fact that antitrust cases, by nature, often require an unusually high level of very costly factual and economic analysis and present specific difficulties for claimants when it comes to access to crucial pieces of evidence often kept secret in the hands of the defendants, deter many victims from bringing actions as they consider the risk/reward balance to be negative.[41]

48. As the Comparative Study on the conditions of damages claims has shown, legal uncertainty exists at several levels. For a number of important legal issues, existing national law is simply unclear about which rule applies in the specific context of antitrust damages cases.[42] These issues include, to give just one example, the availability of the “passing-on defence”.[43] This means a situation where a company that purchased from an infringer was in a position to reduce its economic loss by passing on the overcharge to its own customers. In most Member States, there are no clear rules and hardly any case law on this crucial issue.

49. Additional legal uncertainty for stakeholders stems from the significant differences between the procedural and substantive rules governing actions for damages in the individual Member States. These are described in the Comparative Study on the conditions of damages claims.[44] For instance, while in some Member States strict liability for damages exists once an infringement of the competition rules has been established, in others victims must produce proof that the infringer acted at least negligently.[45] Another example of big differences between the Member States are the possibilities (or lack thereof) for victims to gain access to, and use in court, evidence which is in the possession of the opponent. Also the way in which the damage is quantified differs considerably between Member States, for instance the treatment of losses of (business) opportunities or of interest. Such diversity in the legal systems is not only a problem for victims, be they consumers or businesses, when they try to bring an action for damages in another Member State, e.g. at the domicile of the defendant. For businesses, the potential of becoming a defendant in a damages action in another Member State under national rules which are unfamiliar to them and even unclear in objective terms, creates imponderability and costs that can have a negative impact on the functioning of the internal market. Moreover, in certain circumstances companies in Member States with less effective antitrust damages rules and procedures enjoy an undue competitive advantage in terms of potential liability over companies from Member States with more effective rules.[46] All in all, it is therefore clear that a more level playing field should be in the interest of both potential victims and companies doing business across the EU.

50. Uncertainty amongst stakeholders about the applicable rules even exists with regard to the established acquis communautaire on damages for harm in the field of competition, which is mainly derived from judgments on the Treaty rules by the European Court of Justice. Victims of infringements of the competition legislation (and their legal advisers) are often unaware that the Court has spelled out a number of significant practical consequences that flow from the general principles governing protection of the Community law entitlement to damages.[47] These examples include certain specific requirements concerning the beginning, duration and suspension of limitation periods under national law which determine the time after which a claim for damages can no longer be brought. They also include specific implications, following from the Community law principle of effectiveness, for the national rules on standing ( locus standi ) in damages cases, i.e. the procedural conditions that determine which group of persons is entitled to start a lawsuit for damages. Also in the case of the types of damage that victims must be able to recover under national law for breach of EC competition law, the acquis communautaire establishes certain principles[48] not well known to victims.

51. As mentioned earlier, alongside the legal uncertainty described, the national rules on civil law and civil procedure contain a series of direct difficulties and obstacles for victims seeking effective redress in competition cases. These have been set out and analysed in more detail in the Green Paper and the accompanying Staff Working Paper. During the formal and informal consultation, many stakeholders confirmed their existence and importance. This group of obstacles in the rules of civil procedure and civil law comprises the following aspects:

52. Obtaining access to the evidence necessary for proving a case for damages is considered to be one of the main problems faced by victims for a combination of two reasons: competition cases are extremely fact-intensive and require assessment of a variety of very complex factual elements, while it is often very difficult, or even impossible, for claimants to produce the required evidence, since it is in the possession of the defendants. Many Member States apply strict rules requiring claimants to proffer precisely identified pieces of evidence to support all the facts of their case and impose only limited obligations on defendants to reveal crucial evidence in their possession. This makes it extremely difficult for victims to overcome the information asymmetry present in most competition cases.[49]

53. The damage that antitrust infringements cause to consumers and small businesses is often scattered between a large number of victims. In view of the existing procedural difficulties and the complex analysis required in antitrust cases, such victims are unlikely to bring individual damages actions given the costs of the litigation in relation to the size of their individual claims.[50] However, few Member States currently allow effective collective redress mechanisms resulting in the award of damages.[51] This particularly affects SMEs and final consumers.

54. As mentioned earlier, the lack in many Member States of clear written rules or even guidance from case law concerning the passing-on defence creates significant uncertainty about the outcome of actions and deters potential claimants from taking the potentially high cost risk of an action.

55. Where a private action for damages follows a decision by the European Commission, existing legislation (Article 16 of Regulation 1/2003) provides that claimants for damages can rely on this decision as unrebuttable evidence of the infringement before civil courts. In the majority of Member States, decisions adopted by the national competition authorities (NCAs) lack any such binding effect. [52] Only in a few Member States can claimants rely before civil courts on such public decisions as unrebuttable evidence of the infringement found. In the others, defendants are allowed to call into question the findings of the NCA decision and the civil courts presented with a claim for damages can be required to re-examine the facts already investigated and established by the specialised NCA, even if its decision was confirmed by appeal courts. Even if such relitigation as regards the finding of the infringement would in most cases be unsuccessful, it would add a further imponderability and, especially, time and costs to the claimant’s lawsuit and deters victims from bringing an action for damages.

56. Unduly short limitation periods can bar potential claimants from bringing an action.[53] The Member States’ limitation periods for damages actions differ in length and in terms of the starting event (occurrence of damage or knowledge thereof).[54] Particularly in follow-on actions, it can be efficient for parties to wait for the outcome of the public proceedings (and a possible appeal), but in certain instances the rules on limitation periods in Member States do not allow this.

57. In some Member States, claimants who have established that the defendant infringed antitrust rules and caused damage to them still have to prove either negligence or intent on the part of the infringer.[55] Litigation on the issue of fault adds to the costs and duration of proceedings and increases uncertainty, especially when the relevant factual evidence to determine whether an undertaking applied the requisite standard of care are in the possession of the undertaking.

58. Although the case law of the Court of Justice clarifies several issues regarding the types of damage that are recoverable, significant differences and uncertainty exist in the Member States’ laws as regards the way in which victims can calculate the damage suffered . Quantification can be particularly difficult, since the claimant has to produce evidence of a hypothetical scenario, namely what his economic situation would have been in the absence of the infringement.[56] In many Member States it is unclear to what extent judges can have recourse to approximate methods of calculating damages, such as the ex aequo et bono method or the use of rebuttable presumptions, in cases where exact calculation of the damage proves practically impossible or excessively difficult.[57]

59. The last, but by no means least, driver of the current ineffectiveness of damages actions relates to the costs of actions and the rules on their allocation . The risk of excessive costs (due to the difficulties of bringing a case for antitrust damages under the current legal framework) or of overly unpredictable costs (due to the legal imponderability described and the difficulty of foreseeing the complex economic assessments in many antitrust cases) amplifies the chilling effect of the other obstacles described.[58]

Need for EU action in view of the likely evolution of the problem

60. The problems described above and the overall ineffectiveness of the legal framework for antitrust damages actions are not new. Member States have been required to guarantee the effectiveness of EC competition rules since the EC Treaty entered into force. In 2001 the Court of Justice explicitly recalled that, in the absence of Community rules, this includes the obligation for Member States to provide for effective remedies under their rules of civil law and procedure to safeguard the right of all victims to compensation for harm suffered as a result of antitrust infringements. Nonetheless, to date little progress has been made in the vast majority of Member States.

61. Since the Commission’s Green Paper, there has been some increase in reports of attempts by victims to claim damages before civil courts, some of which were successful.[59] Besides greater awareness on the part of victims of their entitlement to damages, as confirmed by the Court of Justice, this increase in the number of lawsuits appears to be partly due to certain legislative developments in some Member States[60] and also to the emergence of new strategies designed to circumvent some of the traditional obstacles in antitrust damages cases.[61] However, monitoring of developments by the Commission and the findings of the Impact Study[62] show that successful damages actions are still scarce, that in many Member States the problems identified in the Green Paper remain essentially unchanged and that there is, therefore, still a clear need to improve the conditions for effective compensation of victims for harm suffered as a result of infringements of competition law.

62. As mentioned earlier, in recent years a small number of Member States have enacted legislation aimed at facilitating antitrust damages cases.[63] However, such isolated initiatives by Member States cannot be expected to be capable of effectively and fully remedying the root causes of the problem. First, there is no indication that any sizeable number of other Member States are likely to introduce, in the foreseeable future, legislative changes that will ensure an effective legal framework for damages actions brought by victims of antitrust infringements. Second, by nature, isolated initiatives by Member States cannot ensure that a consistent minimum level of effective protection of victims’ entitlement to damages under Articles 81 and 82 will be achieved in every Member State. Nor will they be able to provide a more level playing field for businesses or to reduce the uncertainty created by the current big differences between the national legal systems. Third, even the Member States which have completed the most comprehensive reforms have not addressed all the issues identified in the Green Paper and the accompanying Staff Working Paper.[64] Finally, the interaction between measures facilitating private enforcement of competition rules and various aspects of public enforcement needs to be addressed, for instance the operation and protection of the Commission’s and Member States’ leniency programmes. Individual action by Member States does not seem capable of achieving this in any consistent manner.

63. These issues will be addressed further in the discussion of the principles of subsidiarity and proportionality in section 6.3, as the relevant considerations and the evaluation depend very much on the specifics of the policy option chosen. Nonetheless, at this stage it can already be noted that any solution that is to remedy the main aspects of the problem identified effectively will, most likely, require some form of action coordinated at EU level.

Objectives − the “WHAT”

64. This section sets out the general policy objective pursued, along with several more specific underlying objectives. Based on these objectives, section 5.1 sets out and explains a set of specific assessment criteria that make it possible to measure, in qualitative terms, to what extent the various policy options considered are capable of contributing to achieving the general and specific objectives pursued.

General objective

65. The general objective of this policy initiative is to ensure that victims of infringements of EC competition law have access to truly effective mechanisms for obtaining full compensation for the harm they suffered.

66. Victims of infringements of EC competition law have a right to compensation conferred by the EC Treaty. However, to date, as underlined in section 2, the current legal framework in which actions for antitrust damages are brought makes it considerably difficult for victims to enforce their right. By pursuing this objective, the Commission wishes to guarantee, in every Member State, certain minimum common standards allowing victims effectively to claim, and obtain, full compensation from the infringers of the EC competition rules.

67. Today, the often very remote possibility of having to compensate for the harm caused by infringements of the EC competition rules poses no credible risk to wrongdoers. Rendering the right to compensation more effective will therefore create additional deterrence. Effective legal remedies will also enhance the incentives for infringers, when caught, to settle out of court with the victims and, consequently, improve the quality of settlements for the victims.

68. Achieving this objective of more effective compensation will ensure that the costs of infringements of competition law are borne by the infringers, and not by the victims, by compliant businesses and, indirectly, by society as a whole. This is in line with the competitiveness objectives for the EU.

69. Since the primary objective pursued is full compensation of victims, the damages to be awarded should not influence the level of fines imposed by competition authorities in their public enforcement activities, nor under any future framework of enhanced private actions. Public fines and purely compensatory damages serve two distinct objectives that are complementary: the main objective of public fines (and of potential criminal sanctions) is to deter not only the undertakings concerned (specific deterrence) but also other undertakings (general deterrence) from engaging or persisting in behaviour contrary to Articles 81 and 82. The main objective of private damages is to foster corrective justice by repairing harm caused to individuals or businesses. Of course, as mentioned earlier, this by no means precludes that effective systems for provision of damages also have positive side-effects on deterrence.

70. Given these different functions and primary objectives, in all Member States the level of civil damages (to the extent that they are intended to compensate victims) is independent of whether or not any public fines (or criminal sanctions) have been or are likely to be imposed, and the level of public fines is independent of whether or not any civil damages have been or are likely to be awarded.

71. Whilst courts are specialised and equipped to adjudicate in individual disputes between private parties, e.g. over payment of compensation, public enforcement authorities, by contrast, are not in a position to award damages to victims. They are neither empowered nor equipped to investigate and calculate the harm suffered by each individual victim and to determine the amount to be distributed to individuals as compensation. The task of competition authorities is to focus on the public interest, not on the interest of individuals. Requiring competition authorities to do the latter, would pose a risk of blurring their role and function.

72. In view of the foregoing, the Commission considers that enhanced private actions cannot and should not be intended to replace public enforcement and that public enforcement cannot and should not replace awards of civil compensation to victims.

Specific objectives

73. Given the variety of factors which contribute to the difficulties which victims have in obtaining compensation, the general objective of full compensation can be broken down into more specific objectives. These specific objectives will allow a more systematic and thorough assessment of whether the general objective is fulfilled.

74. Full compensation

75. Ensuring full compensation for the entire harm suffered: attainment of this objective requires no under-compensation of any victim. In particular, the damage awards should include pre-judgment interest in order to compensate the victims for the real value of the harm suffered.

76. While avoiding over-compensation: measures put forward as a result of this initiative should not lead to victims receiving damages higher than the entire loss suffered.

77. Greater awareness of the rules and deterrence, increased enforcement and improved compliance, to the benefit of Europe’s competitiveness

78. Improving compliance with the rules and deterrence: EC competition rules are a matter of public order and greater compliance will be achieved by rendering more credible the risk that infringers will have to compensate the victims and, in particular, by increasing the probability both of civil suits being brought and of infringements being detected.

79. Increasing victims’ awareness of their entitlement to damages and of the conditions for bringing a claim to court.

80. Increasing (potential) wrongdoers’ awareness of the rules governing actions for damages and clarifying the conditions for their liability.

81. Reinforcing European competitiveness by means of greater compliance with the EC competition rules.

82. At the same time, avoiding that the policy leads to over-deterrence, where the risk of claims for damages prevents undertakings from engaging in lawful conduct or where damages are to be paid by undertakings which have engaged in lawful conduct or where such companies face high costs to defend themselves against unmeritorious claims. The principle of proportionality must be observed.

83. Access to justice

84. Guaranteeing effective access to the courts for all victims, including those with scattered low-value damages, such as consumers.

85. Ensuring that potentially high costs do not deter victims from bringing their claims.

86. Facilitating access to the relevant facts and, hence, the truth in a case, in particular overcoming information asymmetry in terms of evidence.

87. Putting in place the conditions ensuring that meritorious claims can effectively be brought, and be successful, despite the complexity of antitrust damages cases.

88. At the same time, putting in place the conditions to avoid abusive litigation and ensuring that unmeritorious claims do not lead to the award of damages.

89. Appropriate and efficient use of the judicial system

90. Streamlining handling of antitrust damages cases by the courts by means of economies of scale, in particular by joining or grouping identical or similar claims.

91. Reducing the costs of litigation by improving the conditions for settlements: settlements are cost-efficient and, when possible, are to be preferred to court actions. However, as settlements are voluntary, attainment of this objective presupposes the existence of a credible court alternative if no settlement is reached. The credible court alternative will also serve as a benchmark leading to improvement of the quality of the settlements.

92. Limiting procedural abuses: while the victims should have better access to the courts, it is important that law-abiding undertakings do not bear the costs and burden of abusive litigation. It is therefore necessary to have appropriate safeguards to prevent such abuses.

93. Limiting the risk of multiple litigation on identical or similar issues: relitigation of issues already settled would entail unnecessary costs and delays plus the risks of a diverging outcome. As far as possible, relitigation should therefore be avoided.

94. A more level playing field and increased legal certainty for businesses operating throughout Europe

95. Ensuring a more level playing field in Europe: attainment of this objective will ensure that businesses across Europe compete on an equal footing and that European citizens and undertakings can enforce their rights conferred by the EC Treaty equally throughout Europe.

96. Increasing legal certainty for businesses: businesses operating throughout Europe will benefit from increased legal certainty as a result of setting minimum common standards for their liability for infringements of competition law.

97. At the same time, respecting national legal traditions and values: a more level playing field should not be achieved without taking due account of the national legal systems and the balance struck over time by each Member State in its national rules.

Policy options − the “HOW”

Identifying the policy options to be assessed

98. In order to determine whether, and if so to what extent, it is appropriate to recommend specific measures in the White Paper, the Commission considered the benefits and costs of a wide range of potential policy options. These are not only specific options identified in the Commission’s Green Paper but also options that emerged during the subsequent consultations and discussions. The Commission did not consider it appropriate, at the stage of preparing a White Paper, to concentrate on one narrow policy approach and to analyse only variations thereon in the impact assessment. Instead, the Commission believes that it is important to assess the full spectrum of possible options, ranging from introduction in Europe of legislative measures that would provide for a system similar to the one applied in countries with highly developed private enforcement of competition rules creating incentives at one end to the zero option of taking no EU action at all (be it legislative or non-regulatory) at the other.

99. Given the diverse causes for the current ineffectiveness of antitrust damages actions, the policy options assessed in this report consist of bundles of several specific measures that address, to different degrees, individual issues such as access to evidence, collective redress, passing-on, costs, etc. These bundles of specific measures are referred to as “Policy Options”.

Method for identifying and grouping the individual measures into bundles to form Policy Options

100. Each Policy Option (i.e. bundle) and its constituent parts reflect a common policy line, a similar degree of stringency and a comparable level of regulatory intervention. In other words, individual measures from the wide spectrum have been grouped together, with the result that the Policy Options analysed themselves range over a wide spectrum of possible policy lines and levels of regulatory intervention. To illustrate this, at one end of the spectrum a Policy Option was selected that pursues the policy of creating maximum incentives for claimants to bring damages actions by means of far-reaching legislative changes to the existing civil law systems. Towards the other end, another Policy Option aims at facilitating damages actions to a lesser degree and envisages achieving this by purely non-regulatory means. The most conservative Policy Option is, obviously, “no action at EU level”.

101. In the first step , five Policy Options − each comprising a set of individual measures each pursuing a comparable policy line and relying on a similar level of regulatory intervention − are assessed (see section 5). Assessment of these bundles of measures by no means precludes that, during this impact analysis, individual measures within these bundles might be identified as appropriate for inclusion in a new bundle that would ultimately become the Preferred Option. Indeed, in this case the balance of positive and negative impacts of the five Policy Options initially analysed did lead the Commission to the conclusion that one of these Policy Options is preferable to the others, but still shows room for improvement, in the form of adjusting some of its individual measures in order to obtain the desired results most effectively at the lowest possible cost (see sections 6.1.1 and 6.1.2). In the second step , the Preferred Option identified in this way is then subjected to the same full impact assessment as the five Policy Options initially analysed. This will allow an overall comparison and illustration of the impact of each of the options analysed. The Preferred Option is therefore subjected to the same steps of impact assessment as the other options. The two-step analysis is therefore primarily a question of presentation and order of analysis. The purpose is to show that the Commission has assessed bundles of measures pursuing a common policy line and similar level of regulatory intervention but is also open to consider a combination of measures from different bundles.

Individual measures not included in the Policy Options analysed

102. Although not every individual measure considered in the Green Paper could be included in the Policy Options analysed in this report, all those measures which received strong support from stakeholders during the consultation process are covered. The specific measures were selected and grouped into a Policy Option first on the basis of the detailed qualitative assessment of individual measures contained in the Impact Study and the results of the consultation process and, second, with a view to achieving a consistent overall policy line within each Policy Option.[65] Specific measures that are incompatible with Community law in the light of existing case law were discarded from the Policy Options at an early stage. This primarily concerned the issue of legal standing to sue.[66] In this context, some advocate concentrating damages actions on the level of direct purchasers (i.e. the direct customers of the company infringing competition law) and denying indirect purchasers (customers further down the distribution chain) the right to sue for any harm they may have suffered (indirectly) as a result of an antitrust infringement. In this respect, the Court of Justice has made it clear that “any individual” must have the right to seek “compensation for the harm suffered where there is a causal relationship between that harm and an agreement or practice prohibited under Article 81 EC”.[67] As a result, the option of excluding indirect purchasers from standing to sue has not been included.

103. Another possibility considered was to discard from the outset, for reasons of legal compatibility, inclusion of multiple damages in any of the Policy Options. Multiple (punitive) damages (as opposed to purely compensatory damages) raise serious issues as regards their compatibility with the public policy and/or basic principles of tort law in many Member States.[68] Under Community law, the existence of exemplary or punitive damages in Member States may be acceptable as the Court clarified in its Manfredi judgment that “in accordance with the principle of equivalence, it must be possible to award particular damages, such as exemplary or punitive damages, pursuant to actions founded on the Community competition rules, if such damages may be awarded pursuant to similar actions founded on domestic law”[69] (however, this does not imply that such particular damages should be introduced in every Member State). Therefore, with a view to subjecting the full spectrum of possible (and sometimes supported) solutions to an impact assessment, it was decided not to discard a priori double damages from the Policy Options, without ignoring that in some Member States there are legal objections to punitive damages. Particular attention was therefore paid to assessing the feasibility under national law and the impact of such measures (see the tables in section 5 under the heading “implementation costs”).

104. Finally, it should be recalled that introducing higher fines for infringers or criminal sanctions on executives involved are no means of achieving the primary objective of this policy, namely more effective compensation of victims. Such measures may act as a stronger deterrent for potential infringers and, thereby, have the positive effect of limiting the number of victims suffering from breaches of EC competition law. However, they will contribute nothing to repairing the harm suffered by victims of (undeterred) antitrust infringements, who will remain uncompensated in the absence of measures enhancing the effectiveness of private actions for damages such as those proposed in the various Policy Options.

Measures regarding interaction with public enforcement

105. Finally, due to the policy consideration of avoiding possible negative consequences for public enforcement as a result of any system of enhanced actions for damages, all the Policy Options (except the zero action option) contain the same two measures in this respect: protection from disclosure of corporate statements made in the context of leniency programmes and limitation of civil liability on the part of successful immunity applicants to their direct or indirect contractual partners. As a result of the latter measure, the only persons entitled to receive compensation from the immunity recipient would be the victims who directly bought the cartelised products or services from the immunity recipient (“direct contractual partners”) and those further down the supply chain who bought the same products or services (directly or via intermediaries) from the direct contractual partners (“indirect contractual partners”). As a consequence, the immunity recipient would not be held liable for the damages attributable to cartelised products or services sold by co-cartelists.

106. Other forms of impact that some measures may have on public enforcement are taken into account wherever they are likely to lead to appreciable costs, in particular, litigation costs and administrative burden for competition authorities (see also section 5.3).

Provision of non-binding assistance for quantification of damages

107. The Policy Options described below relate to possible measures to be included in the White Paper. It should be noted that, alongside the White Paper, the Commission is also considering providing pragmatic, non-binding assistance in the difficult task (often on economic grounds) of quantifying damages in antitrust cases, for the benefit of both national courts and the parties.

Description of the Policy Options assessed

108. Based on the considerations set out above, the Commission focused on the Policy Options described below as the most likely potential policy choices in view of the current ineffectiveness of antitrust damages actions. At one end of the spectrum, Policy Option 1 envisages legislative measures providing for maximum facilitation and incentives for damages claims while, at the other, Policy Option 5 envisages the scenario of zero EU action, either legislative or non-regulatory. Table 1 provides an overview of all five Policy Options and their main components.

Policy Option 1: legislative measures maximising facilitation of claims and incentives for victims

109. Policy Option 1 envisages a set of legislative measures[70] that would not only remove almost all[71] the recognised obstacles to effective antitrust damages actions but also introduce strong incentives for claimants to bring an action. This approach is characteristic of countries which rely strongly on damages actions to perform a deterrent function in the public interest. To set such incentives, Policy Option 1 groups together measures that include, first, the availability of double damages for all types of infringement of Articles 81 and 82. As damages, in a European context, are understood as also covering payment of interest accrued since the damage occurred,[72] double damages can be as much as three times as high as damages in countries where no pre-judgment interest is awarded.[73] Introducing double damages for every type of infringement would not only create an incentive for victims to bring a claim, but also contain a punitive element aimed at increasing deterrence of infringers.

110. Beyond that, Policy Option 1 provides for “opt-out” class actions[74] allowing a small number of victims to sue and claim damages for the entire group they belong to, except for victims who decide not to be included in the class. It would also introduce mandatory one-way fee-shifting, so that the claimant would never be liable for the defendant’s costs, regardless whether the claimant loses or wins the case, unless the claim was frivolous or vexatious.

111. The rules on disclosure would provide for very broad access to evidence: the claimant would be allowed to request any relevant document from the defendant once he has passed the relatively low threshold of notice pleading.[75] Concerning the issue of passing-on and claims by indirect purchasers, the passing-on defence would be excluded in order to provide an additional incentive for direct purchasers to sue, while indirect purchasers would nonetheless, for legal reasons (see above), still be allowed to claim damages if they can show that they have suffered harm caused by the infringement.

112. In civil actions following a decision by a competition authority, Policy Option 1 would avoid costly and time-consuming relitigation by rendering the findings of a national competition authority in the EU (if not appealed against or if confirmed on appeal) binding on civil courts.[76] Similarly, Policy Option 1 would provide for strict liability once an infringement of Article 81 or 82 has been established. This means that in these cases claimants would not have to prove separately negligence or intent on the part of the infringer. Claims for antitrust damage would be time-barred only after a relatively long limitation period of twenty years from the occurrence of the harm or, if the claimant could reasonably have been aware of his right to compensation, five years as of that moment.

113. With the aim of avoiding the risk that enhanced damages actions might have a negative impact on public enforcement activities and, in particular, the functioning of leniency programmes, Policy Option 1 suggests two measures: protection from disclosure of corporate statements made in the context of leniency programmes and limitation of civil liability on the part of the successful immunity applicant to his direct and indirect contractual partners.

Policy Option 2: substantial facilitation and partial incentives for damages claims by means of legislative measures

114. Policy Option 2 envisages a range of legislative measures[77] addressing the recognised obstacles to effective antitrust damages actions in a less far-reaching/radical manner than Policy Option 1. Also, in terms of encouraging claimants to bring an action, Policy Option 2 differs from Policy Option 1 in that it offers a much smaller number of incentives: double damages would be introduced only for cartels as they are hardcore infringements, whereas for other types of infringement damages would be limited to compensation for the full harm suffered (including pre-judgment interest). The rule that the losing party pays the costs, prevalent in the EU, would remain, but be made subject to the courts’ discretionary power to shift costs to the defendant, e.g. where − in meritorious cases − there is a marked asymmetry in the financial resources of the parties.

115. The collective action mechanism under Policy Option 2 would not take the form of an “opt-out” class action, but, instead, of “opt-in” collective action, where only victims who have decided to join the group seeking compensation are included in the action. This collective redress mechanism would be supplemented by representative action, where a qualified representative body, such as a consumer association or trade association, claims damages for the harm caused to the interests of those it represents.

116. Disclosure of documents would be broad and follow an initial exchange between the parties of lists indicating all relevant evidence, but only after the claimant has presented reasonably available facts and evidence in support of the allegations (fact pleading). With regard to the passing-on defence and claims by indirect purchasers, both offensive and defensive use of the passing-on argument by claimants and defendants would be permitted. In the case of claims by indirect purchasers, a risk allocation rule would apply, facilitating proof of passing-on for the downstream victim. As regards the requirement in several countries’ national law that fault on the part of the infringer needs to be proven by the victim, Policy Option 2 would introduce a presumption of fault that can be rebutted if the infringer can show that the breach of competition rules was due to an excusable error.

117. As in Policy Option 1, decisions of all national competition authorities would have binding effect across the EU. The limitation period would be set at five years from the date on which the claimant should reasonably have been aware of the harm (subjective standard). If public proceedings are initiated, a new limitation period of two years would begin on conclusion of the proceedings (including any appeals).[78] With the aim of avoiding the risk that enhanced damages actions might have a negative impact on public enforcement activities and, in particular, the functioning of leniency programmes, Policy Option 2 also suggests the same two measures: protection from disclosure of corporate statements made in the context of leniency programmes and limitation of civil liability on the part of the successful immunity applicant to his direct and indirect contractual partners.

Policy Option 3: moderate facilitation by means of more limited legislative measures

118. Policy Option 3 is aimed at addressing the most significant obstacles to an effective system of damages claims by means of legislative measures[79] of more limited scope, but not at offering further incentives for claimants to bring a case. Under Policy Option 3, only single damages would therefore be available, with pre-judgment interest. Once claimants have passed the threshold of fact pleading (see paragraph 84) that can reasonably be required in the circumstances, access to evidence in the hands of the other party would be available. To avoid overly broad disclosure, claimants would have to identify specific categories of evidence to be disclosed and courts would monitor the proportionality of the request for disclosure.

119. As regards the requirement of fault, Policy Option 3 would introduce a rule that establishment of the infringement strongly suggests the existence of at least negligence on the part of the infringer.

120. As in Policy Option 2, both claimants and defendants could fully avail themselves of the passing-on argument. Discretionary fee-shifting by the court would be possible, although only for a portion of the legal costs.[80] Decisions of national competition authorities would be binding only on the courts of the Member State whose authority issued the decision. In terms of collective redress, only representative actions would become available. The limitation period under Policy Option 3 would be five years from the date on which the claimant should reasonably have been aware of the harm. In the event of public proceedings, the limitation period would be suspended pending a final decision and any appeals.[81] With the aim of avoiding the risk that enhanced damages actions might have a negative impact on public enforcement activities and, in particular, the functioning of leniency programmes, Policy Option 3 suggests two measures: protection from disclosure of corporate statements made in the context of leniency programmes and limitation of civil liability on the part of the successful immunity applicant to his direct and indirect contractual partners.

Policy Option 4: moderate facilitation by means of non-regulatory measures

121. Policy Option 4 can be dubbed the “non-regulatory approach” as it entails no legislative measures at EU level. Instead, this option would identify, mainly from the experience in Member States, a range of useful solutions and best practices, which would be recommended to all Member States for consideration for implementation in their own legal systems.

122. Under Policy Option 4, the same individual rules (e.g. on collective redress, passing-on and access to evidence) would be envisaged as under Policy Option 3, although not as legislative measures at Community level, but as non-binding recommendations. Policy Option 4 therefore relies, to a large extent, on the Member States to carry out the action suggested.

Policy Option 5: zero action at EU level (baseline option)

123. The fifth and final option − as the baseline option − entails taking no action at all at EU level with regard to antitrust damages actions. It is possible that further developments could arise as a result of Community action in other areas, such as consumer protection. Unilateral action by Member States in this or other areas, along with interpretations of Community law by the Court of Justice, could also affect the status quo with regard to damages claims. For the foreseeable future, however, the Commission would refrain from proposing any legislation or non-regulatory measures specific to antitrust damages actions in Europe. The assessment of the impact of Policy Option 5 examines the status quo and any developments considered likely without any EU action.

Table 1: Overview of the Policy Options analysed

Access to evidence | Broad disclosure, low threshold | Initial provision of lists + broad disclosure, both based on fact-pleading threshold | Disclosure of specified categories, based on fact-pleading and proportionality |

Indirect purchaser | Standing allowed | Standing allowed | Standing allowed |

Passing-on | Defence excluded | Defence allowed; facilitation of proof of pass-on in favour of indirect purchaser | Defence allowed |

Effect of NCA decisions | Binding across EU | Binding across EU | Binding in own Member State |

Fault (once infringement established) | Strict liability | Rebuttable presumption; exoneration for excusable errors | Strong probative value of finding of infringement |

Collective redress | Opt-out class actions | Opt-in collective + representative actions | Representative actions |

Limitation period | 20 years as of damage or 5 years as of reasonable knowledge | Minimum 5 years as of reasonable knowledge + restart + two years | Minimum 5 years as of reasonable knowledge + suspension |

Cost rule | One-way shifting | Loser pays, but judge may shift all costs | Loser pays, but judge may shift part of costs |

Interaction with leniency | Protection of corporate statements from disclosure; limitation of liability on the part of successful immunity applicant | Protection of corporate statements from disclosure; limitation of liability on the part of successful immunity applicant | Protection of corporate statements from disclosure; limitation of liability on the part of successful immunity applicant |

ANALYSIS OF IMPACT

Assessment criteria

124. A multi-criteria analysis was used to assess and compare the impact: each Policy Option was assessed against a set of criteria relating to different potential benefits and costs. These are explained in more detail below (see section 5.1.1 for potential benefits and section 5.1.2 for potential costs). Because of the non-availability of statistics, it is not possible to quantify the likely impact of each Policy Option in monetary terms. Therefore, for each Policy Option, the impact expected in respect of a range of potential benefits and costs has been assessed in qualitative terms.

Benefits

a. Ensuring full compensation of the entire harm suffered

125. Though an important assessment criterion, the degree to which this specific objective is considered achieved will largely coincide with the degree to which a given Policy Option allows victims of an infringement of competition law to bring an action and enables them to prove the infringement, the harm and the causal link between the two. Once those hurdles are cleared, little room for manoeuvre is left as regards the scope of the damages award, in the light of the Manfredi judgment of the Court of Justice.

126. According to this judgment, “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.”[82] Any award of damages that would not fully compensate for the entire harm falling within the scope of this judgment would therefore violate the principle of effectiveness.

127. Since the objective is full compensation for the entire harm suffered, Policy Options will score lower on this criterion to the extent that they are likely to lead to under- or over-compensation of the claimant. Claimants are under- or over-compensated when they receive less or more damages than the harm actually suffered. The latter is a fortiori the case if the claimant receives damages but has suffered no harm whatsoever. Since full compensation in the greatest possible number of cases is the first and foremost objective, good scores on achieving the goal of full compensation will weigh heavily in the impact analysis and in identification of the Preferred Option.

b. Increased awareness, enforcement, deterrence and legal certainty

128. Policy Options will score higher to the extent that they make economic operators more aware of their rights and obligations under EC competition rules. Clear and explicitly formulated rules add to such awareness, just as much as they clarify the conditions applicable to claims for damages and the conditions of liability on the part of offenders.

129. Likewise, Policy Options will score higher to the extent that they widen the scope or increase the: intensity of enforcement (by simply increasing the number of cases for which infringers are held responsible, by addressing different types of infringement or by involving a wider variety of economic operators); likelihood of detection and of having to bear the financial consequences of anti-competitive behaviour; and, assuming that the optimum level has not yet been reached, the level of those financial consequences.

130. A higher degree of awareness of the EC competition rules, combined with better enforcement of them (by means of private actions), may contribute to greater compliance with these rules and, hence, to better achievement of their objectives. Any impact inherent in more effective compensation of victims obviously has to be taken into account as a potential benefit, given that more effective deterrence and greater compliance with the rules will contribute to achieving the results of fair competition.

131. Policy Options will therefore score higher if they are likely to lead, all other things being equal, to an increase in deterrence rates. However, because increasing deterrence is not one of the primary objectives, in the final stage of comparing Policy Options and determining the Preferred Option less weight will be given to positive scores on deterrence, particularly compared with good scores on the central objective of compensation.

132. Policy Options will score lower on this category to the extent that they lead to a situation where the risk of damages claims prevents undertakings from engaging in lawful conduct or where damages are to be paid by undertakings that did not infringe the competition rules.

c. Access to justice

133. Policy Options will score higher on this criterion to the extent that they ensure more effective access to justice for all victims of an infringement of competition law. That is particularly relevant for victims who have only small claims and/or have difficulties in gaining access to the evidence necessary to prove their case. Policy Options will score well if they adequately address the reasons why these victims decide not to bring a damages case, although they would like to claim the damages they are entitled to.

134. At the root of these reasons is the fact that victims fear the opportunity cost and the financial consequences of losing a meritorious case. More than the former factor, which is largely a matter of personal judgment, the latter is the key to ensuring adequate access to justice, e.g. by reducing the financial risks or the likelihood that meritorious cases will be lost. Consequently, Policy Options that, for instance, allow effective and efficient collective redress and facilitate access to evidence and provide for accuracy in fact-finding will score well on the “access to justice” criterion.

d. Efficient use of the judicial system

135. Efficient use of the judicial system means avoiding unnecessary delays, multiple proceedings and contradictory outcomes. Policy Options will therefore score higher to the extent that they allow related cases to be dealt with together. Cases are considered related when they concern the same infringement, whereas identity of claimant and/or defendant is not really required. Key issues raised by the case, like establishment of the infringement, damage calculation, etc., may indeed be similar, independent of the litigating parties. Even if some of these issues vary between the parties, claims concerning the same infringement are likely to be interrelated and would therefore benefit from being handled jointly.

136. As the judicial system should not be overburdened, Policy Options will, paradoxically, also score higher to the extent that they allow cases to be settled adequately out of court.

137. Policy Options will score lower on this criterion to the extent that they offer claimants excessive incentives. Incentives can be considered excessive if they encourage claimants to bring damages claims although they have suffered no harm whatsoever. Policy Options that offer incentives to claimants who suffered only minor harm are not considered to encourage procedural abuses and therefore do not score lower on this criterion for that reason alone, even if the harm is outweighed by the litigation costs that go with the claim.

e. A more level playing field

138. Having similar rules for antitrust damages claims throughout Europe is considered beneficial, as it leads to equal treatment for businesses and consumers who are in a similar situation. Particularly for businesses, exposure to antitrust damages claims may have an influence on their market behaviour. Fair competition on the internal market would therefore require comparable exposure to damages claims, which can be brought about only by similar procedural rules governing actions for damages.

139. Since antitrust damages actions are put forward by the Court of Justice as a remedy stemming directly from European law, claimants of antitrust damages should all be able to use this remedy effectively. Although the effectiveness of the remedy may allow some divergence between Member States, it does require compliance with certain minimum standards. Those minimum standards would have to be the same across Europe.

140. As a result, Policy Options will score higher to the extent that they create a level playing field in Europe for claimants and defendants in antitrust damages actions.

Costs

a. Litigation costs

141. This broad category of costs covers both the litigation costs for parties to proceedings (both settlement costs and costs incurred when the case is brought to court) and the enforcement costs for public authorities (such as courts and competition authorities). Policy Options will therefore score higher on these costs to the extent that they offer incentives to litigate and/or suggest measures that increase the costs (resources, opportunity cost and money) of litigation for the parties or for public authorities.

b. Administrative burden

142. In line with the European Commission’s Impact Assessment Guidelines, this category includes costs incurred by businesses, consumers and public authorities in order to meet legal obligations to keep information and to provide it, either to public authorities or to private parties. Only the net costs are taken into account, i.e. excluding those that would be incurred anyhow, even without any legal obligation.

143. Policy Options will score higher on this cost to the extent that they require businesses, consumers or public authorities to keep information for a long period of time (storage costs) and that they impose an obligation to provide that information to one of the parties (disclosure costs). Policy Options will score high on disclosure costs to the extent that they have a low threshold triggering disclosure and/or wide scope of disclosure.

c. Error costs

144. Error costs means costs related to the possibility of courts issuing a mistaken decision. That could take the form either of incorrectly awarding damages (type I error) or of incorrectly rejecting a claim for damages (type II error). Policy Options will score high on error costs to the extent that they suggest measures that increase the likelihood of error and/or measures that amplify the impact of the error.

145. Errors in quantification of damages are not included in this category because the resulting under- or over-compensation already has a negative impact on the corrective justice objective (see point 95).

d. Implementation costs

146. Implementation costs means costs incurred by businesses, consumers and national public authorities to adapt to new rules (e.g. training costs, compliance costs, etc.). Although real, these costs are therefore transitory. Policy Options will score high on these costs to the extent that they lead to a big change in the regulatory context in which businesses, consumers and public authorities operate.

Other impacts

147. Apart from these categories of costs and benefits, this report also assesses the impact of the different Policy Options on consumers and SMEs on the one hand and on macro-economic variables, such as competitiveness, innovation, growth and jobs, on the other. The latter impact coincides largely with the expected level of compliance with the competition rules. In particular, the more that undertakings comply with the rules, the more competitive markets will be and the lower any allocative inefficiency. Policy Options that are more likely to achieve that result are therefore more likely to contribute positively to growth and employment. Such positive overall effects are likely to outweigh certain negative effects in those rare cases where the breach of competition law and the resulting public fines and liability for civil damages pose a financial threat to the survival of the infringing firm.

Identifying and assessing the impact of each Policy Option

148. As explained in section 1.2.5, the assessment presented in this report is based on the Commission departments’ own analysis, the results of a broad consultation of stakeholders and, in particular, the findings of the Impact Study. This section sets out, in the form of tables, the conclusions of the Commission’s assessment of the positive and negative impacts that Policy Options 1 to 5 would be likely to have if implemented.

149. The point of reference, for the purposes of this assessment, is the absolute impact of the various Policy Options. In other words, each option is assessed on its own merits, including Policy Option 5, which is the “no policy change” or “business as usual” (BAU) scenario. The reason for choosing this method is that, in the context of claims for damages for breach of the EC antitrust rules, it is important to illustrate the significant costs and benefits that taking no action at all (“business as usual”) would have. This approach will, in particular, make it possible to compare all Policy Options with each other, including Policy Option 5.

Policy Option 1

150. The legislative measures envisaged under Policy Option 1 would aim at maximum facilitation of damages claims and offering very strong incentives for victims. Central components of this Policy Option include double damages for all types of infringement, “opt-out” class actions, mandatory one-way fee-shifting, a rule excluding the passing-on defence, very broad disclosure of evidence at a relatively low threshold, binding effect of all NCA decisions, strict liability and a potentially long limitation period.

151. Tables 2 to 4 set out the main conclusions of the Commission’s analysis of the likely positive and negative impacts of Policy Option 1.

Table 2 − Benefits of Policy Option 1

Benefits achieved/ problem addressed | Impact: zero (0) to high (((((() | Explanation of rating and aspects of the policy option most relevant to the benefits |

1. Ensuring full compensation of the entire harm suffered | (((( | Overall, most effective for obtaining damages, but high risk of over-compensation: opt-out class actions and the various incentives lead to a large increase in number of victims compensated • double damages create likelihood that (at least) full harm is compensated, even if cases are settled at a lower amount than claimed • more accurate fact-finding (very broad disclosure) allows more precise quantification of loss suffered and increases number of victims compensated • very long limitation period makes compensation for entire duration of infringement more likely • one-way fee-shifting can allow victims to settle for an amount that is closer to the actual harm; • but risk that double damages might lead to over-compensation, although this risk is mitigated in settlements • in opt-out class actions, possible difficulties in identifying victims and distributing damages compensation may be averaged • in opt-out class actions, the very large group of victims included in the class may not always be able to control the lawyers acting for the class (principal/agent problem) misalignment of their interests could lead to under-compensation •absence of passing-on defence may lead to over-compensation of direct purchasers • limitation of liability on the part of successful immunity applicants may reduce available compensation (if other infringers are insolvent), although this risk is limited.[83] |

2. Increased awareness, enforcement, deterrence and legal certainty | (((( | Overall, enforcement, deterrence and legal certainty highly increased, but risk of over-deterrence in non-cartel cases: double damages: opt-out class actions increase potential liability of infringers per case and in total • no over-deterrence in cartel cases, inter alia as the detection rate for cartels is likely to remain very low (but improved, although due to secrecy less than for other types of infringement), i.e. not higher than 20% • rules enforced much more frequently, increasing deterrence: the various incentives may stimulate monitoring and detection by market players and consumers • opt-out class actions necessitate publication of claims and that raises awareness• exclusion of passing-on defence increases the number of claims by direct purchasers and their chances of success• comprehensive clarification of conditions for exercising rights and liability • but significant risk of over-deterrence in abuse and vertical cases, although somewhat mitigated by settlements • high risk of strategic/unmeritorious lawsuits.[84] |

3. Access to justice | ((((( | Overall, very broad access to justice, especially for indirect purchasers: high incentives and opt-out class actions will lead to compensation of very much higher number of victims (including those who suffered scattered damages)• greatly facilitated access to evidence and greater accuracy in fact-finding• legal costs are removed as a hurdle to access to justice due to one-way fee-shifting• binding effect of NCA decisions/strict liability facilitate proof significantly • access to justice for longer time due to limitation periods.[85] |

4. Efficient use of the judicial system | ((( | Overall improvement in efficiency, but significant potential for abuses and excesses: binding effect of NCA decisions and strict (no-fault) liability clearly improve efficiency in follow-on cases • absence of pass-on defence simplifies proof/calculation of damage • opt-out class actions simplify small-claims actions • but, especially, very broad discovery at low threshold and automatic fee-shifting create high risk of strategic lawsuits • unmeritorious cases impose significant burden on courts.[86] |

5. A more level playing field | ((((( | Equal protection of right to damages across the EU a practically level playing field for consumers and businesses alike. |

Table 3 − Costs of Policy Option 1

Costs | Costs: zero (0) to high (((((() | Explanation of rating and aspects of the policy option most relevant to the costs |

1. Litigation costs | ((((( | Overall, very high increase in total litigation costs and high increase per case: opt-out class actions, double damages and one-way fee-shifting highly increase number of lawsuits • opt-out class actions and very broad discovery at low threshold make cases costly, although opt-out class actions are more efficient than individual suits • strong risk of unmeritorious litigation • abuses such as “discovery blackmail” can impose significant burdens on undertakings • limitation of liability on the part of successful immunity applicants may increase litigation costs for victims, although for the defendants some costs are likely to arise even without this measure from the additional litigation by the cartelist first held liable who will seek a contribution from the co-infringers • but binding effect of NCA decisions reduces parties’ costs in follow-on claims • early disclosure/strict liability reduce costs during trial • early disclosure and limitation of liability on the part of successful immunity applicants may stimulate cost-efficient early settlements.[87] |

2. Administrative burden | ((((( | High increase in administrative burden: more frequent litigation/early and very broad discovery at low threshold/long limitation period mean that a much greater number of companies must screen and produce many more documents more often over (in many MS) a longer period (until end of limitation period).[88] |

3. Error costs | ((( | Overall, costs of errors likely to rise (few errors with high impact): double damages, fee-shifting and opt-out class actions imply high cost impact of each error • in addition, the increased magnitude of potential errors might sometimes induce judges wrongly to reject claims in close cases (to avoid the enormous negative impact of a wrong finding of liability) • but greater accuracy in fact-finding binding effect of NCA decisions/strict liability may reduce errors in follow-on cases.[89] |

4. Implementation costs | ((((( | A very costly change for all MS, with major legal hurdles for several (especially as regards double damages): various measures under this option require very significant changes to the law of all MS, which implies a very significant need for training of judges and the legal community • double damages, opt-out class actions, very broad discovery rules at low threshold and absence of passing-on defence raise public policy or even constitutional concerns in several MS, which may be difficult to overcome • early (automatic) disclosure at low threshold without judicial involvement clashes with conception of civil procedure in many MS with a civil law tradition • limitation of liability on the part of successful immunity applicants may be difficult to implement in various MS (e.g. because in a cartel the contribution by each participant is crucial for the cartel’s success and can therefore be causal for the damage suffered by every victim) • although many of these are one-off costs, they are nonetheless extremely high.[90] |

Table 4 − Other impacts of Policy Option 1

Other impacts | Benefit: zero (0) to high (((((() | Explanation of rating and aspects of the policy option most relevant in this context |

1. Positive impact on SMEs and consumers | ((( | SMEs and consumers, often with limited financial resources and relatively risk-averse, benefit most from the incentives and facilitation of damages claims under this Policy Option • but SMEs and consumers might also suffer from increased litigation activity, especially if upstream players can, depending on market conditions, pass on their increased costs.[91] |

2. Likely macro-economic impact | (( | Greater enforcement and resulting deterrence would, in principle, push for more competitive markets with likely positive effects on growth and employment • but risk of over-deterrence and risk of high costs/heavy burden and of strategic lawsuits might lead to less efficient resource allocation and R&D activity and negative consequences on prices.[92] |

Policy Option 2

152. The legislative measures envisaged under Policy Option 2 aim at addressing almost all the recognised obstacles to effective antitrust damages actions, although in a less far-reaching manner than Policy Option 1. Policy Option 2 also offers a smaller number of incentives. The main features of this Policy Option include double damages for hardcore cartels, a “loser pays” cost rule coupled with a discretionary power for courts to shift all costs to the defendant, “opt-in” collective and representative actions, broad disclosure of evidence based on fact-pleading and an initial exchange of lists of relevant evidence, a rule allowing the passing-on defence and facilitating claims by indirect purchasers, a presumption of fault that can be rebutted in the event of an excusable error, binding effect of all NCA decisions and a restart of the limitation period on conclusion of the public proceedings.

153. Tables 5 to 7 set out the main conclusions of the Commission’s analysis of the likely positive and negative impacts of Policy Option 2.

Table 5 − Benefits of Policy Option 2

Benefits achieved/ problem addressed | Impact: zero (0) to high (((((() | Explanation of rating and aspects of the policy option most relevant to the benefits |

1. Ensuring full compensation of the entire harm suffered | (((( | Overall, very effective for obtaining damages, but some risk of over-compensation: the various measures to facilitate actions and incentives should lead to a clear increase in the number of victims compensated • double damages for cartels create a likelihood that (at least) full harm is compensated, even if cases settle at a lower amount than claimed • greater accuracy of fact-finding (broad and early disclosure, subject to fact-pleading) allows more precise quantification of loss suffered • facilitation, for indirect purchasers, of proof of passing-on makes compensation of such victims more likely • opt-in collective/representative actions make recovery of scattered damage more likely and allow better alignment of compensation with actual harm than Policy Option 1 • non-exclusion of passing-on defence is in line with the compensation objective • five-year limitation plus restart normally sufficient to allow recovery • but some risk that double damages for cartels might lead to over-compensation, although risk mitigated in settlements; •some risk of under-compensation where non-cartel cases are settled (at a lower amount than actual harm) • limitation of liability on the part of successful immunity applicants may reduce available compensation (if other infringers are insolvent), although this risk is limited.[93] |

2. Increased awareness, enforcement, deterrence and legal certainty | ((((( | Overall, significant increase in deterrence, enforcement and legal certainty: opt-in collective/representative actions, double damages for cartels and binding effect of NCA decisions plus presumption of fault increase likelihood of (private) enforcement and potential liability of infringers (in each case and in total) and, thereby, deterrence • no risk of over-deterrence • stimulation of monitoring by market players and consumers and greater accuracy of fact-finding likely to lead to greater detection rates (although likely to remain very low for cartels) • comprehensive clarification of conditions for exercise of right to damages and for liability of companies to occur • rule on limitation period allows effective compensation in most cases while avoiding over-long legal uncertainty.[94] |

3. Access to justice | ((((( | Overall, access to justice is broader, especially for indirect purchasers: compensation of greater number of victims (including those who suffered scattered damages) • opt-in collective/representative actions improve access to justice for low-value small claims • facilitation of pass-on makes proof of damage more likely • early access to evidence facilitated • discretionary full fee-shifting can encourage victims to file meritorious but low-probability suits • the rule on the limitation period, especially the restart, allows proper access to justice • greatly facilitated access to evidence and greater accuracy in fact-finding • binding effect of NCA decisions/presumption of fault facilitate proof significantly.[95] |

4. Efficient use of the judicial system | ((((( | Overall, very significant improvement in efficiency: binding effect of NCA decisions and fault presumption enhance efficiency in follow-on cases • early discovery list can shorten proceedings • opt-in collective/representative actions allow more efficient aggregation of individual claims.[96] |

5. A more level playing field | (((( | Similarly effective protection of right to damages across the EU: more level playing field for consumers and businesses alike. |

Table 6 − Costs of Policy Option 2

Costs | Costs: zero (0) to high (((((() | Explanation of rating and aspects of the policy option most relevant to the costs |

1. Litigation costs | ((( | Overall, increase in costs in total and per average case: clear increase in number of law suits • broader disclosure than status quo in many MS can increase litigation costs • disclosure rule entails some risk of “fishing expeditions” and “discovery blackmail” as judicial control occurs only ex post (although risk limited by the possibility of discretionary fee-shifting and fact-pleading threshold) •opt-in collective /representative actions may entail new costs for courts • limitation of liability on the part of successful immunity applicants may increase litigation costs for victims, although for the defendants some costs are likely to arise even without this measure from the additional litigation by the cartelist first held liable who will seek a contribution from the co-infringers • but opt-in collective/representative actions produce efficiencies that can reduce costs per case compared with individual actions • binding effect of NCA decisions/presumption of fault reduce parties’ costs in follow-on claims • early disclosure and limitation of liability on the part of successful immunity applicants may stimulate cost-efficient early settlements • binding effect of NCA decisions across the EU facilitates concentration of damages claims in multi-state cases in one court.[97] |

2. Administrative burden | ((( | Overall, a significant impact: increase in number of lawsuits, broader disclosure than in most MS and automatic early drafting of lists in every lawsuit mean that a greater number of companies must screen and produce more documents more often • in some MS longer record-keeping obligations due to longer limitation period than currently • but the burden is clearly lighter than under Policy Option 1.[98] |

3. Error costs | (( | Number of errors may increase in total, as the number of cases increases due to incentives and facilitation of damages actions • double damages for cartels imply a certain cost impact of each error in cartel cases • in addition, the increased cost impact of potential errors might induce judges wrongly to reject claims in close cartel cases • but the statistical incidence of errors would be reduced •binding effect of NCA decisions may make errors less likely in follow-on cases;• greater accuracy in fact-finding • the “loser pays” rule (as a general principle) stimulates selection of meritorious cases and prevents frivolous suits.[99] |

4. Implementation costs | (((( | Very significant implementation costs for many MS and major legal hurdles for several MS (as regards double damages): various measures under this option require significant changes to the law of many MS in a few MS little change required • significant need for training of judges and the legal community • double damages for cartels raise public policy concerns in several MS which may be difficult to overcome • early (automatic) disclosure of evidence without prior judicial involvement does not fit easily into the conception and basic structure of civil procedure in many MS with a civil law tradition • limitation of liability on the part of successful immunity applicants may be difficult to implement in various MS (e.g. because, in a cartel, the contribution by each participant is crucial for the cartel’s success and can therefore be causal for the damage suffered by every victim) • although all these are one-off costs, they are nonetheless very significant.[100] |

Table 7 − Other impacts of Policy Option 2

Other impacts | Benefit: zero (0) to high (((((() | Explanation of rating and aspects of the policy option most relevant in this context |

1. Positive impact on SMEs and consumers | ((( | SMEs and consumers are likely to benefit most from the incentives and facilitation of damages claims under this policy option • but slight risk that the “loser pays” rule (fee-shifting only optional) might discourage SMEs and consumers in meritorious low-probability cases.[101] |

2. Likely macro-economic impact | ((( | Effective second pillar of enforcement and resulting deterrence would, in principle, push for more competitive markets with likely positive effects on growth and employment • much lower risk than in Policy Option 1 of excessive litigation leading to a deteriorating business environment.[102] |

Policy Option 3

154. Policy Option 3 envisages moderate facilitation by means of more limited legislative measures. Its main features include single damages with pre-judgment interest for all infringements, disclosure of specified categories of evidence based on fact-pleading and judicial control of proportionality, the possibility of passing-on and indirect purchaser claims, discretionary fee-shifting by courts for part of the legal costs, binding effect of NCA decisions in the same Member State, only representative actions in terms of collective redress and suspension of the limitation period while public proceedings are in progress.

155. Tables 8 to 10 set out the main conclusions of the Commission’s analysis of the likely positive and negative impacts of Policy Option 3.

Table 8 − Benefits of Policy Option 3

Benefits achieved/ problem addressed | Impact: zero (0) to high (((((() | Explanation of rating and aspects of the policy option most relevant to the benefits |

1. Ensuring full compensation of the entire harm suffered | ((( | Overall, a more effective compensation mechanism, but not all victims benefit: clear but moderate facilitation of damages claims (especially disclosure of categories of evidence, representative actions and discretionary partial fee-shifting) is likely to lead to some increase in the number of victims compensated • in many MS, disclosure of categories of evidence will improve the likelihood of proving liability and quantification of the full actual harm suffered • no risk of over-compensation • binding force of domestic NCA decisions and suspension of limitation period during public investigations improves chances of (efficient) compensation in follow-on actions • non-exclusion of passing-on defence is in line with the compensation objective • indirect purchasers can claim passed-on damages • but absence of any facilitation to prove pass-on keeps these cases very difficult for indirect purchasers • collective redress mechanism (representative only) less appropriate for certain types of cases: qualified entities may focus on priority cases • risk of under-compensation (especially in the case of a cartel) where cases are settled at a lower amount than the actual harm • limitation of liability on the part of successful immunity applicants may reduce available compensation (if other infringers are insolvent), although this risk is limited.[103] |

2. Increased awareness, enforcement, deterrence and legal certainty | ((( | Overall, moderate increase in deterrence, enforcement and legal certainty: some increase in magnitude of damages award, especially via representative actions, with disclosure rules leading to a moderate increase in deterrence • some increase in detection rate, although still very low for cartels • no risk of over-deterrence • significant clarification of conditions for exercise of right to damages and for liability of companies to occur • but some risk of under-deterrence for cartels.[104] |

3. Access to justice | ((( | Overall clearly improved, but insufficient, especially for victims with relatively low-value claims: more victims gain access to justice than in the status quo • in terms of access to evidence, significant improvement in several MS, as disclosure of categories of evidence possible and as initial fact-pleading threshold is adapted to circumstances of each case • discretionary partial fee-shifting alleviates cost barrier for claimants to some extent, depending on how much judges can/will shift costs in practice • but collective redress mechanism (representative only) less appropriate for certain types of cases: qualified entities may focus on priority cases • mere suspension of limitation during public investigation can prove problematic in some follow-on cases • purely domestic binding effect of NCA decisions does not support concentration of damages claims in multi-state cases in one court.[105] |

4. Efficient use of the judicial system | ((( | Overall, some improvement of efficiency: in some scenarios, representative actions allow efficient aggregation of small claims • binding effect of domestic NCA decisions enhances efficiency in follow-on cases (only in national cases) • no significant risk of increase in unmeritorious litigation as the “loser pays” rule has a certain case-selection effect (while optional fee-shifting maintains access to justice) • no disclosure abuses because ex ante judicial control, especially of proportionality • but absence of collective actions makes proceedings less efficient for small claims • increased burden on courts compared with status quo in several MS • in multi-state cases, the rule on the bindingeffect of domestic NCA decisions would probably produce the less efficient result that victims bring a separate action in each MS that adopted a public decision separately.[106] |

5. A more level playing field | ((( | Minimum level of protection of right to damages in all MS: more level playing field for consumers and businesses alike • but significant differences between MS remain as many MS would maintain at least some rules that go further than the minimum level of protection of the victims’ right (especially on the issues of collective redress, fault requirement and passing on). |

Table 9 − Costs of Policy Option 3

Costs | Costs: zero (0) to high (((((() | Explanation of rating and aspects of the policy option most relevant to the costs |

1. Litigation costs | (( | Overall, litigation costs increase moderately in total and decrease moderately per average case: some increase in number of lawsuits (less than the number of victims compensated due to representative actions) • disclosure subject to fact-pleading and ex ante proportionality test plus the “loser pays” principle (subject to discretionary, partial fee-shifting) lead to moderate increase in costs and a very limited risk of procedural abuses, if any • costs for claimants and for courts per case may even decrease due to efficiencies of representative actions (compared with individual actions), even though for certain types of cases representative actions are less appropriate than collective actions • in some MS (where such a disclosure rule is currently uncommon), increase in the burden on courts • purely domestic binding effect of NCA decisions increases costs as it does not support concentration of damages claims in multi-state follow-on cases in one court •limitation of liability on the part of successful immunity applicants may stimulate cost-efficient early settlements, but may also increase litigation costs for victims, although for the defendants some costs are likely to arise even without this measure from the additional litigation by the cartelist first held liable who will seek a contribution from the co-infringers.[107] |

2. Administrative burden | (( | Overall, a relatively small impact: slight increase in number of lawsuits and broader disclosure than in several MS lead to slightly more screening and production of documents • in some MS longer record-keeping obligations due to longer limitation period than currently • but the administrative burden is lighter than under Policy Option 2.[108] |

3. Error costs | ( | Number of errors may increase slightly in total, if at all: as the number of cases rises slightly, so may the total number of errors, but there is no indication that the overall statistical incidence of errors would increase • beneficial impact of NCA decisions on errors limited by the absence of cross-border binding effect • but greater accuracy in fact-finding due to access to greater amount of evidence in several MS • the “loser pays” rule stimulates appropriate selection of cases and prevents frivolous suits.[109] |

4. Implementation costs | (( | Moderate implementation costs: some measures under this option (especially disclosure rules and representative actions, but also the limitation periods and binding effect) require changes to the law of several MS • need for training of judges and the legal community • none of the changes required raises major public policy concerns • limitation of liability on the part of successful immunity applicants may be difficult to implement in various MS (e.g. because in a cartel the contribution by each participant is crucial for the cartel’s success and can therefore be causal for the damage suffered by every victim) • all these are one-off costs.[110] |

Table 10 − Other impacts of Policy Option 3

Other impacts | Benefit: zero (0) to high (((((() | Explanation of rating and aspects of the policy option most relevant in this context |

1. Positive impact on SMEs and consumers | (( | Consumers benefit most from the introduction of representative actions and possibility of partial fee-shifting as their interests are represented at a lower cost • but SMEs benefit to a lesser extent as some representative entities are more likely to focus on consumer harm, may face significant limitations in terms of resources in some MS and, in certain cases, non-consumer/commercial associations may not be able to act due to conflicting interests (e.g. where the infringers are a crucial supplier or are members of the same trade association in a given sector) • still a risk that the “loser pays” rule (partial fee-shifting being only optional) and strict disclosure criteria might discourage SMEs and consumers, especially if they do not possess much evidence at the outset.[111] |

2. Likely macro-economic impact | (( | Less likely that a very effective second pillar of enforcement will emerge, therefore weaker push for more competitive markets likely to have positive effects on growth and employment • very low risk of excessive litigation leading to a deteriorating business environment.[112] |

Policy Option 4

156. Policy Option 4 is the “non-regulatory approach” as it would identify, mainly from the experience in Member States, a range of useful solutions and best practices, which the Commission would recommend to all Member States for implementation in their own legal systems. The impact assessment of the non-regulatory approach of Policy Option 4 is based on the assumption that the same specific measures would be recommended for Member State action as those suggested in Policy Option 3 as legislative measures at EU level.

157. Predicting the impact of mere recommendations such as those in Policy Option 4 is a particularly delicate exercise. In practice, it is difficult to foresee the degree to which Member States will actually act on the Commission’s recommendations and, where needed, adopt the corresponding new rules at national level. The likelihood of such legislative changes is particularly doubtful as it would imply amending sometimes long-standing rules in an area of law (civil law and procedure) which is generally not known for frequent intervention by the legislature.

158. The likelihood of Member States acting on the recommendations and, hence, the likelihood of Policy Option 4 producing the desired benefits of improving the effectiveness of damages actions would, besides a more general “awareness-raising” effect, seem to depend mainly on what is called a “moral suasion” effect.

159. There is a stream of literature advocating use of “soft law” to achieve broad policy objectives at international and EU levels. Benchmarking, such as recommendation of best practices, has been advocated, in particular, by international organisations to promote new modes of governance and better regulatory practices. However, there is no compelling evidence that the “moral suasion” effect has ever been strong enough to create a high likelihood that EU Member States would change their existing national laws[113], particularly not in “long-standing” areas of law such as civil liability and civil procedure. There is therefore no guarantee that the “moral suasion” effect of a recommendation will lead all, or even the majority of, Member States to produce the desired results, namely an actual improvement in protection of victims’ right to compensation.

160. The rating of various impacts of Policy Option 4 set out below nonetheless reflects both hypotheses, i.e. both effective and rather ineffective “moral suasion”. The rating is therefore expressed as a range indicating, at the upper end, the likely impact under the hypothesis that all Member States act on the recommendations and introduce the rules and best practices suggested. At the lower end, the rating indicates the likely impact under the hypothesis that, at the very best, a few Member States follow the recommendations. In the case of the lower end, this is still a rather conservative (i.e. optimistic, but not very realistic) assumption, because it may well be that no Member State follows the recommendations. Under these hypotheses, the lower end of the range corresponds largely with the likely impacts of Policy Option 5 and the upper end with the likely impacts of Policy Option 3.

161. On this basis, Tables 11 to 13 set out the main conclusions of the Commission’s analysis of the likely positive and negative impacts of Policy Option 4.

Table 11: Benefits of Policy Option 4

Benefits achieved/ problem addressed | Impact: zero (0) to high (((((() | Explanation of rating and aspects of the policy option most relevant to the benefits |

1. Ensuring full compensation of the entire harm suffered | ( − ((( | The likely impacts of Policy Option 4 correspond, under the hypotheses defined above, at the upper end of the range with the likely impacts of Policy Option 3 and, at the lower end, with the likely impacts of Policy Option 5. Consequently, see the explanations in the tables assessing Policy Options 3 and 5 for the range of ratings in this table. |

2. Increased awareness, enforcement, deterrence and legal certainty | ( − ((( | See the explanations in the tables assessing Policy Options 3 and 5 for the range of ratings in this table. |

3. Access to justice | ( − ((( | See the explanations in the tables assessing Policy Options 3 and 5 for the range of ratings in this table. |

4. Efficient use of the judicial system | ( − ((( | See the explanations in the tables assessing Policy Options 3 and 5 for the range of ratings in this table. |

5. A more level playing field | 0 − ((( | See the explanations in the tables assessing Policy Options 3 and 5 for the range of ratings in this table. The differences may even widen if some MS act on the recommendations and others maintain the status quo. |

Table 12 − Costs of Policy Option 4

Costs | Costs: zero (0) to high (((((() | Explanation of rating and aspects of the policy option most relevant to the costs |

1. Litigation costs | ( − (( | See the explanations in the tables assessing Policy Options 3 and 5 for the range of ratings in this table. |

2. Administrative burden | 0 − (( | See the explanations in the tables assessing Policy Options 3 and 5 for the range of ratings in this table. |

3. Error costs | ( | See the explanations in the tables assessing Policy Options 3 and 5 for the ratings in this table. |

4. Implementation costs | 0 − (( | See the explanations in the tables assessing Policy Options 3 and 5 for the range of ratings in this table. |

Table 13 − Other impacts of Policy Option 4

Other impacts | Benefit: zero (0) to high (((((() | Explanation of rating and aspects of the policy option most relevant in this context |

1. Positive impact on SMEs and consumers | ( − (( | See the explanations in the tables assessing Policy Options 3 and 5 for the range of ratings in this table. |

2. Likely macro-economic impact | 0 − (( | See the explanations in the tables assessing Policy Options 3 and 5 for the range of ratings in this table. |

Policy Option 5

162. Policy Option 5 entails zero action at EU level (i.e. neither legislation nor non-regulatory measures) on antitrust damages actions in Europe. The assessment of the impacts of Policy Option 5 therefore examines the status quo and any developments considered likely without EU action. The likelihood of any such developments is difficult to foresee, even for the short and medium term. It should be borne in mind that Member States have been required to guarantee the effectiveness of the EC competition rules since the EC Treaty entered into force. In 2001 the Court of Justice explicitly called upon Member States to provide, in the absence of Community rules, for effective remedies under their civil law and procedure to safeguard the right of all victims to compensation for harm suffered as a result of infringements of Articles 81 or 82 of the EC Treaty. Notwithstanding this, to date very few Member States have taken action to improve the effectiveness of their legal framework for antitrust damages actions and there is no clear indication that a significant number of other Member States would adopt the necessary measures without any impetus from EU level.

163. The Commission’s analysis of various likely impacts of the option of zero action at EU level was based on the probably still conservative (i.e. optimistic) assumption that some Member States would adopt a number of measures to improve the effectiveness of antitrust damages actions in Europe. Tables 14 to 16 set out the main conclusions of the Commission’s analysis of the likely positive and negative impacts of Policy Option 5.

Table 14 − Benefits of Policy Option 5

Benefits achieved/ problem addressed | Impact: zero (0) to high (((((() | Explanation of rating and aspects of the policy option most relevant to the benefits |

1. Ensuring full compensation of the entire harm suffered | ( | Under-compensation of many victims likely to remain: no evidence that in the absence of Community action the number of victims compensated would increase significantly • obstacles to bringing actions and to successfully proving the conditions for compensation highly likely to persist in many MS • consequently, no guarantee that all EU citizens will enjoy a certain minimum level of protection of their right to antitrust damages.[114] |

2. Increased awareness, enforcement, deterrence and legal certainty | ( | Continued under-deterrence, especially for cartels and lack of legal certainty: overall, at best, very small increase in deterrence rates and compliance by means of private actions for damages • clear under-deterrence, especially for cartels, in most MS • uncertainty resulting from the differences between the national legal systems will persist • no increase of awareness at pan-European level, at best in some MS • without any guarantee of a minimum level of protection, victims of antitrust infringements would have no clear picture of their basic right under Community law to damages in the MS • the lack of a common approach to the interaction with public enforcement (e.g. protection of corporate statements by leniency applicants) creates a serious risk that inconsistencies or even loopholes might hamper public enforcement.[115] |

3. Access to justice | ( | In most MS, access to justice, in particular for SMEs and consumers, remains problematic: even though some measures improving collective redress may well be taken in a few MS, access to justice in antitrust damages cases will continue to be difficult, as in many MS significant changes are unlikely, especially in the following areas: access to evidence, cost rules, limitation periods and their suspension/restart, effect of NCA decisions and the problem of legal uncertainty, e.g. in connection with pass-on.[116] |

4. Efficient use of the judicial system | ( | Current inefficiencies remain in most MS, except maybe some improvement as a result of better collective redress in certain MS.[117] |

5. A more level playing field | 0 | Substantial differences remain between MS: highly unlikely that in the absence of any EU action businesses would compete across Europe on an equal footing • highly unlikely that citizens and undertakings can enforce their rights conferred by the Treaty throughout the EU in an equal/similar manner • no indication that a virtuous “mutual learning” process or a “race to the top” in the form of competition between legal orders would stimulate adoption of similar best practices across the EU • the current fragmentation of the legal framework for damages actions could even become wider in a few years if some MS enact significant reforms while others maintain the status quo. |

Table 15 − Costs of Policy Option 5

Costs | Costs: zero (0) to high (((((() | Explanation of rating and aspects of the policy option most relevant to the costs |

1. Litigation costs | ( | Litigation costs may be significant in individual cases, but relatively low overall and total costs might increase slightly: exercising the right to damages is costly (especially due to lack of widespread effective collective redress, legal uncertainty, difficult access to accurate evidence, lack of binding force and differences between legal systems), but the level of litigation is low • number of cases may increase slightly, if some MS facilitate damages actions • but if certain MS improve collective redress, costs for claimants may decrease as a result of efficiencies (compared with individual actions).[118] |

2. Administrative burden | 0 | Relatively small burden in most MS, but may increase slightly: relatively small number of companies concerned due to low number of cases • in most MS, currently limited information/disclosure obligations on other party • in the not very likely event that some MS increase the scope of disclosure and limitation periods in antitrust damages cases, a somewhat heavier administrative burden may result.[119] |

3. Error costs | ( | Little litigation some errors will persist, but total number moderate: some errors likely due to lack of effective access to evidence and, hence, access truth and absence of learning effects • given low level of litigation, total number of errors not high • but some MS might possibly introduce rules for greater accuracy in fact-finding (although not very likely) • to the extent that the number of cases increases, courts will grow more familiar with antitrust cases and avoid errors.[120] |

4. Implementation costs | 0 | No implementation costs: in the absence of any provisions at EU level, no costs to implement them.[121] |

Table 16 − Other impacts of Policy Option 5

Other impacts | Benefit: zero (0) to high (((((() | Explanation of rating and aspects of the policy option most relevant in this context |

1. Positive impact on SMEs and consumers | ( | See comments on access to justice above: exercise of right to damages would remain very difficult: consumers and SMEs are also likely to suffer particularly from the persisting legal uncertainty • some improvement could occur in the area of collective redress in some MS.[122] |

2. Likely macro-economic impact | 0 | Negligible contribution.[123] |

Likely costs for public authorities and businesses under the Policy Options

164. In its analysis of the likely positive and negative impacts of the various policy options, the Commission paid particular attention to the costs that each measure envisaged would imply for businesses on the one hand and public authorities on the other. These various types of cost are indicated in Tables 2 to 16.

165. Costs for businesses can occur, with differences between the Policy Options, in all four categories analysed in detail above. Particularly relevant to undertakings are “litigation costs”, “administrative burden” and “error costs”, but in some instances “implementation costs” may also be of relevance.

166. The courts of Member States would, naturally, play an essential role under any Policy Option to improve the effectiveness of damages actions. The potential costs and burden in this context are accounted for in the above tables under the headings “litigation costs” and “efficient use of the judicial system”. Costs to the legislature, judiciary and administrations as a result of the need, in certain Policy Options, to adopt and implement amendments to the existing laws are counted as “implementation costs”.

167. In the assessment and comparison of the various Policy Options in section 6, the impact on businesses and public authorities in terms of these costs plays an essential role.

Other impacts, in particular likely implications for the EU budget

168. In principle, this policy initiative has no direct impact on the EU budget. No new measures for training judges (or similar measures) are envisaged in addition to those already ongoing and provided for in the EU budget. Nor are special information campaigns at EU level envisaged as part of the policy. The only budgetary implication which the policy will have is the funding required (approximately €200 000) to launch a study to help to prepare for non-binding assistance to judges/parties for quantification of damages (see section 4.1 in fine ).

169. There are no environmental impact of this initiative foreseeable. Trade with, and investments by, non-EU countries will not be affected by the initiative in any appreciable manner. On the contrary, a clearer and more consistent framework for antitrust damages actions across the EU will improve legal certainty and predictability and thereby contribute to a better business environment and make entry into European markets easier, to the ultimate benefit of European consumers.

COMPARING THE OPTIONS

The Preferred Option

Interim results of the impact analysis

170. Based on the impacts analysed above, interim conclusions can be drawn on the strengths and weaknesses of the individual Policy Options and their ability to achieve the objectives properly (especially the objective of full compensation), while avoiding excessive costs.

Policy Option 1

171. Policy Option 1 has great potential for achieving the various objectives.[124] It is particularly efficient at ensuring a more level playing field on the internal market and access to justice, especially for indirect purchasers. It also adequately meets the objective of increasing awareness, enforcement, deterrence and legal certainty, but not fully, due to a risk of over-deterrence, especially in non-cartel cases. The objective of ensuring full compensation of the entire harm suffered is also attained,[125] but, due to a serious risk of over-compensation and certain negative effects of opt-out class actions, this objective is not fully met. This is particularly important since the results in terms of achieving this objective are crucial for the assessment, as full compensation is the objective most directly related to the overall objective of this policy. Furthermore, the costs incurred by this Policy Option are very high. This option has a very significant impact on almost every cost heading, whether litigation costs, administrative burden or implementation costs, especially since it would allow double damages for every type of case and very broad discovery at a low threshold. One particularly important point to note is that this option entails a strong risk of unmeritorious litigation being brought and significant potential for abuses and excesses. Error costs are more limited, due to greater accuracy in fact-finding, even though each error would entail very high costs. As a result, Policy Option 1 does not sufficiently allow proper achievement of the objectives.

172. These findings are in line with the findings of the Impact Study, which, in the context of a largely similar set of measures, found that the costs associated with the measures would be particularly high and would carry the risk of over-deterrence, especially for vertical restraints and abuses of dominance. Also the risk of over-compensation would be tangible due to the double damages envisaged. As a result, the Impact Study concludes that this scenario could lead to development of a litigation culture whose benefits, including in terms of its contribution to macro-economic variables, appear questionable, even compared with the status quo .[126]

Policy Option 2

173. Policy Option 2 contains a range of legislative measures addressing almost all the recognised obstacles to effective antitrust damages actions, although in a less far-reaching/radical manner than Policy Option 1.[127] Its potential for achieving the objectives identified is strong, whatever the specific objective considered.[128] At the same time, it considerably diminishes the risks of over-deterrence and over-compensation observed under Policy Option 1 and, to a very large extent, removes any risk of abuse of the judicial system by unmeritorious claimants. Furthermore, Policy Option 2 is the option that has the most beneficial macro-economic impact. Another consequence of this less far-reaching approach than Policy Option 1 is that the costs incurred under Policy Option 2, though still high, are lower, some slightly (error costs and implementation costs, in particular due to the availability of double damages), others significantly (administrative burden and litigation costs). More specifically, under this option litigation costs would be much lower, in particular due to a more limited increase in the number of cases and a collective action mechanism which would imply lower costs for distributing the damages awarded. Moreover, Policy Option 2 would also pose more limited principal/agent problems than Policy Option 1.[129] Implementation costs are very high, inter alia because the concept of double damages is likely to raise considerable compatibility issues in a number of Member States (see also paragraph 154 below).

174. On balance, Policy Option 2 nonetheless adequately achieves the objectives identified and, at the same time, avoids imposing excessive costs (with the exception of the issue of double damages). This concurs with the findings of the Impact Study which, based on a largely similar set of measures, found that they are likely to contribute positively to both corrective justice and deterrence, while at the same time limiting the risk of proliferation of strategic and abusive litigation.[130]

Policy Option 3

175. Policy Option 3 is aimed at addressing, by means of more limited legislative measures, the most significant obstacles to an effective system of damages claims.[131] In this Policy Option, the costs of achieving the objectives are slightly lower than in Policy Option 2, in particular in the case of implementation costs, but also under every cost heading.[132] Furthermore, the costs of meeting the objectives are significantly lower than in Policy Option 1, in particular the litigation costs, administrative burden and implementation costs.[133]

176. However, one consequence of the individual specific measures contained in this Policy Option is that all the objectives identified in this report are achieved to a lesser extent than under Policy Option 2. This is particularly apparent for the objectives of increasing awareness, enforcement, deterrence and legal certainty, access to justice and efficient use of the judicial system. The objectives are also attained less well than under Policy Option 1, in particular those on access to justice and a more level playing field in Europe. Finally, the positive impact on SMEs and consumers is more limited than under Policy Options 1 and 2. As a result, Policy Option 3 does not sufficiently allow proper achievement of the objectives even though it avoids imposing excessive costs.

177. The analysis carried out in the Impact Study also finds that a similar combination of measures would be likely to prove inadequate, both in terms of achieving more effective damages claims for the individual plaintiffs and for creating a level playing field for businesses and avoiding forum-shopping.[134]

Policy Option 4

178. Policy Option 4 is the non-regulatory option. It provides for no legislative measures at EU level, but proposes identifying a range of useful solutions and good practices at national level to recommend to all Member States for consideration for implementation in their own legal systems.[135] Under this Policy Option, some progress would be made towards achieving the policy objectives, but to a much lesser extent than under Policy Options 1 to 3 (in the best scenario, the objectives would be met as they would under Policy Option 3).[136]

179. Among the various objectives, the one least likely to be achieved is the more level playing field. However, almost all the other objectives would be insufficiently met. In practice, the soft law approach depends on national measures and is likely to result in further fragmentation of the national systems, with Member States applying the solutions and good practices to different extents. On the other hand, the costs of implementing Policy Option 4 are low, the main weakness of this option being its inability to meet the objectives of this policy sufficiently.

180. Essentially, as the findings of the Impact Study confirm, there is little likelihood that the effects of Option 4 would, in practice, add up to much more than those of Option 5, i.e. of inaction at Community level.[137]

Policy Option 5

181. Policy Option 5 is zero action at EU level on this subject. The status quo could be affected by action at Member State level and also by Community action in other areas.[138] However, this option is even less able to achieve the objectives identified than Policy Option 4.[139] Although the costs of implementation are zero, it leads to a situation where error costs cannot be excluded due to the absence of learning effects and lack of sufficient access to accurate evidence. The positive impact on SMEs and consumers would be very limited and the macro-economic impact would be negligible. Overall, it would remain very difficult to exercise the right to damages and the general objective of the initiative would not be met.

182. The unsatisfactory result of the no policy change option is confirmed in the Impact Study, which finds that the level of corrective justice would be very low, while deterrence may increase only very slightly over the next few years. The internal market would remain fragmented in terms of the level of judicial protection and, as a result, prone to forum-shopping. The administrative burden, legislative costs and error costs would be likely to remain at a relatively low level overall. However, significant burdens would be borne by SMEs and consumers who would continue to suffer from inadequate access to justice, while there would be little or no positive macro-economic impact.[140]

Interim conclusion

183. According to this interim analysis of the Policy Options examined so far, Policy Option 2 has the greatest potential for achieving the objectives identified, while avoiding excessive costs.

Identification of the Preferred Option

184. This comparison of the various Policy Options and the characteristics of the underlying specific measures shows that the Policy Option best able to meet the objectives identified effectively at the lowest possible cost is Policy Option 2. This Policy Option meets the general objective of ensuring that victims of infringements of EC competition law have access to truly effective mechanisms for obtaining full compensation for the harm they suffered. Also, it effectively meets all the specific objectives set out in section 3. The costs which would arise should this Policy Option be implemented remain limited, with the highest likely to be the implementation costs, in particular due to the availability of double damages for cartel cases, while error costs would remain very limited. However, this Policy Option comes at a high cost.

185. The analysis of the five Policy Options nevertheless shows that, taking Policy Option 2 as the starting point, potential improvements can be made, in particular by reducing the costs that would be incurred as a result of implementation of this option. To ensure that the most efficient set of measures is proposed in the White Paper, and in view of the impact analysis, the preferred Policy Option (“the Preferred Option”) is a combination of Policy Option 2, with some elements of other less far-reaching Policy Options (3 and 4).

186. The changes compared with Policy Option 2 are in the following areas and for the following reasons:

187. Damages : the analysis of the impact of the various Policy Options justifies switching from Policy Option 2 to Policy Option 3 on the specific measure regarding damages. The change from double damages for cartels to full single damages is more in line with the prevailing principle in most Member States, which is that damages should be of a compensatory nature and should not be construed as an additional sanction (“private penalty”). It must be added, however, that, this may lead to some under-compensation in the event of settlements, especially in cartel cases. The change in the damages rules, from Policy Option 2 to Policy Option 3, actually reduces the costs of the option, be they litigation costs (in particular, as it would reduce the incentive to sue), administrative burden (by decreasing the population subject to each information obligation) or error costs (due to a reduced impact of errors). Moreover, the difference in implementation costs is very significant, since the introduction of double damages proposed under Policy Option 2 would raise serious compatibility issues. Only in a very small number of Member States do civil courts currently have the possibility to award damages that go beyond full compensation of victims and pursue the objective of punishment and deterrence (punitive or exemplary damages). In the other Member States, such punitive damages do not exist. In several Member States, punitive damages are even considered contrary to their legal tradition and the compensatory principle that underlies their law of civil liability. As a result, these Member States, for example, refuse to enforce foreign judgments containing punitive damages on the grounds that they would conflict with the fundamental principles of their own legal order (domestic ordre public /public policy).[141]

188. Access to evidence : the change from Policy Option 2 to Policy Option 3 on the specific measure regarding access to evidence also increases the potential to attain the objectives identified at lower cost. This change means that the rules on access to evidence provide for disclosure of specified categories of evidence, based on fact-pleading and proportionality. As a result, this adjustment of Policy Option 2 implies a certain limitation of the scope of the disclosure rules and the absence of an initial exchange between the parties of lists indicating all evidence relevant to the case. This change has a slight impact on attainment of various objectives. First, the objective of ensuring full compensation of the entire harm suffered will be particularly affected since restricting the scope of disclosure of evidence may make it more difficult to achieve accuracy in compensation. This has a direct impact on the objective of increasing awareness, enforcement, deterrence and legal certainty by slightly lowering the probability of success and due to potential infringers’ reluctance to broader disclosure rules, thereby limiting the deterrent effect of the rule. However, on balance, the rule on access to evidence under Policy Option 3 still allows good achievement of the various objectives identified while significantly reducing the impact on almost every cost heading. First, the change from Policy Option 2 to Policy Option 3 reduces the impact on litigation costs, in particular in the initial litigation phase, allows deeper assessment by the court of the proportionality of the claims and limits the risk of fishing expeditions. Second, this change significantly reduces the administrative burden given the smaller volume of documents to be processed. Third, implementation costs will be lower, as this change will require a smaller number of Member States to amend their national rules regarding access to evidence. The change from Policy Option 2 to Policy Option 3 on the specific issue of access to evidence has no impact on error costs.[142]

189. Cost rule : the change from Policy Option 2 to Policy Option 4 entails no significant change to the substance of the measure regarding the cost rule, but just a change of approach. Under the Preferred Option, the cost rule will not take the form of a legislative measure, but only of identification and recommendation of good practices. Should the good practices be fully implemented at national level, the objectives would be achieved as much as in Policy Option 3, but at lower implementation costs. A non-binding instrument of course entails a strong risk of lack of, or divergence in, implementation at national level, thereby possibly hindering attainment of the objectives. However, given the wide recognition in the Member States of the “loser pays” cost rule, and the fact that in some the court may already deviate from this rule by shifting or capping costs in all or under limited circumstances, the objectives pursued by recommending the cost rule of Policy Option 3 seem to be realistically achievable at national level without legislative intervention at EU level. The incentives to improve the cost rule on the basis of mere identification and recommendation of good practices may be sufficiently strong at national level. As a result, Option 4 should be preferred as regards the cost rule.

190. In summary, the Preferred Option has the following characteristics.

Table 17 − Overview of the Preferred Option

Preferred Option | Policy Option of origin of the measure |

Damages | Full single | Option 3 |

Access to evidence | Disclosure of specified categories, based on fact-pleading and proportionality | Option 3 |

Indirect purchaser | Standing allowed | Options 1 to 3 |

Passing-on | Defence allowed facilitation of proof of pass-on in favour of indirect purchaser | Option 2 |

Effect of NCA decisions | Binding across EU | Option 2 |

Fault (once infringement established) | Rebuttable presumption exoneration for excusable errors | Option 2 |

Collective redress | Opt-in collective + representative actions | Option 2 |

Limitation period | Minimum of 5 years as of reasonable knowledge + restart + 2 years | Option 2 |

Cost rule | No legislative measure, only identification and recommendation of good practices in line with Option 3: loser pays, but judge may shift all or part of the costs | Option 4 |

Interaction with leniency | Protection of corporate statements from disclosure; limitation of liability on the part of successful immunity applicant | Options 1 to 3 |

Impact of the Preferred Option

191. The Preferred Option aims at addressing almost all the recognised obstacles to effective antitrust damages actions. It is highly similar to Policy Option 2, but diverges from it slightly in order to ensure better achievement of the objectives on the one hand and/or limit the implementation costs on the other. The main features of the Preferred Option were described in the previous section. Tables 18 to 20 set out the main conclusions of the analysis of the likely positive and negative impacts of the Preferred Option.

192. In the tables, the Preferred Option is analysed and rated in terms of its likely effects in both absolute and relative terms , i.e. in comparison with the “no policy change” or “business as usual” (BAU) scenario, which was analysed above as Policy Option 5. The reason for these two points of reference is to show the impact of the Preferred Option from every possible perspective and to provide a comprehensive yet concise picture of both the costs and benefits of the Preferred Option in the absolute and compared with the present state of affairs (predicting likely developments in the short to medium term, i.e. 3 to 5 years[143]).

Table 18 − Benefits of the Preferred Option

Benefits achieved/ problem addressed | Impact: zero (0) to high (((((() | Explanation of rating and aspects of the policy option most relevant to the benefits |

absolute | relative to BAU |

1. Ensuring full compensation of the entire harm suffered | (((( | ((( | Overall, a more effective compensation mechanism for all victims: clear but measured facilitation of damages claims (especially by means of disclosure of categories of evidence, opt-in collective and representative actions and, to a lesser extent, discretionary partial fee-shifting if implemented by MS) is likely to lead to an increase in the number of victims compensated • in many MS, disclosure of categories of evidence will improve the likelihood of proving liability and quantification of the full actual harm suffered • no over-compensation • facilitation, for indirect purchasers, of proof of passing-on makes compensation of such victims more likely • non-exclusion of passing-on defence is in line with the compensation objective • binding effect of NCA decisions across the EU allows concentration of damages claims in multi-state cases in one court •availability of both opt-in collective and representative actions make recovery of scattered damage more likely • five-year limitation plus restart normally sufficient to allow recovery • but some risk of under-compensation where cases are settled (at a lower amount than the actual harm) • smaller number of victims bringing claims due to the absence of double damages and more limited disclosure rule • mere recommendations as regards the cost rule pose a strong risk of lack of soft implementation at national level •limitation of liability on the part of successful immunity applicants may reduce available compensation (if other infringers are insolvent), although this risk is limited.[144] |

2. Increased awareness, enforcement, deterrence and legal certainty | (((( | ((( | Overall, increase in deterrence rates, enforcement and legal certainty, but some risk of under-deterrence in cartel cases: increase in magnitude of damages awards, especially by means of opt-in collective and representative actions, plus more effective disclosure rules improve enforcement and, thereby, deterrence rate • some increase in the number of victims compensated adds to this effect • rule on limitation period provides for effective enforcement in most cases while avoiding over-long legal uncertainty •no risk of over-deterrence • significant, comprehensive clarification of conditions for exercising the right to damages and for liability of companies to occur • but no additional stimulation of monitoring and detection by market players by means of double damages • some risk of under-deterrence for cartels.[145] |

3. Access to justice | (((( | ((( | Overall, access to justice is broader, especially for indirect purchasers: compensation of greater number of victims (including those who suffered scattered damages) • opt-in collective actions and representative actions improve access to justice for low-value small claims;• facilitation of pass-on makes proof of damage more likely • in terms of access to evidence, significant improvement in several MS, as disclosure of categories of evidence possible and as initial fact-pleading threshold is adapted to circumstances of each case • discretionary partial fee-shifting, if implemented by MS, can encourage victims to file meritorious but low-probability claims, thereby lowering the cost barrier for claimants to some extent (while maintaining “loser pays” as the basic rule has positive case-selection effects) • rule on limitation period, especially the restart, allows proper access to justice • but disclosure more limited than under Option 2 •mere recommendations as regards the cost rule pose a strong risk of lack of soft implementation at national level.[146] |

4. Efficient use of the judicial system | (((( | ((( | Overall, significant improvement of efficiency: binding effect of NCA decisions and fault presumption enhances efficiency in follow-on cases/proceedings • opt-in collective/representative actions allow efficient aggregation of small claims • no significant risk of increase in unmeritorious litigation as the “loser pays” rule has a certain case-selection effect (while possible discretionary partial fee-shifting maintains access to justice for meritorious but low-probability claims) • no disclosure abuses because ex ante judicial control, especially of proportionality • but mere recommendations as regards the cost rule pose a strong risk of lack of soft implementation at national level.[147] |

5. A more level playing field | (((( | (((( | Similarly effective protection of right to damages across the EU: more level playing field for consumers and businesses alike. |

Table 19 − Costs of the Preferred Option

Costs | Costs: zero (0) to high (((((() | Explanation of rating and aspects of the policy option most relevant to the costs |

abso-lute | rel-ative to BAU |

1. Litigation costs | (( | ( | Overall, moderate increase in total litigation costs and per average case: slight increase in number of lawsuits (less than the number of victims compensated due to collective redress mechanisms) • disclosure subject to fact-pleading and ex ante proportionality test plus “loser pays”/fee-shifting lead to moderate increase in costs and very limited risk of procedural abuses, if any •costs per claimant per case may even decrease due to efficiencies of opt-in collective and representative actions, even though these mechanisms may entail new costs for the courts • in some MS (where such disclosure is currently uncommon), increase in burden on courts • binding effect of NCA decisions across EU allows concentration of damages claims in multi-state cases in one court presumption of fault reduces parties’ costs in follow-on claims • early disclosure may stimulate cost-efficient early settlements • limitation of liability on the part of successful immunity applicants may stimulate cost-efficient early settlements, but may also increase litigation costs for victims, although for the defendants some costs are likely to arise even without this measure from the additional litigation by the cartelist first held liable who will seek a contribution from the co-infringers.[148] |

2. Administrative burden | (( | (( | Overall, a moderate impact: slight increase in number of lawsuits and broader disclosure than in several MS lead to slightly more screening and production of documents • in some MS longer record-keeping obligations due to longer limitation period than currently.[149] |

3. Error costs | ( | Number of errors may increase slightly, if at all: as the number of cases increases slightly, so may the total number of errors, however, no indication that overall statistical incidence of errors would increase •binding effect of NCA decisions may make errors less likely in follow-on cases • greater accuracy in fact-finding • the “loser pays” rule stimulates selection of meritorious cases and prevents frivolous suits.[150] |

4. Implementation costs | (( | (( | Moderate implementation costs: some measures under this option (especially disclosure rules, binding effect and, to a lesser extent, collective redress mechanisms) require changes in the law of several MS • need for training of judges and the legal community • none of the changes required raises major public policy concerns • limitation of liability on the part of successful immunity applicants may be difficult to implement in various MS (e.g. because in a cartel the contribution by each participant is crucial for the cartel’s success and can therefore be causal for the damage suffered by each victim) • these are all one-off costs.[151] |

Table 20 − Other impacts of the Preferred Option

Other impacts | Benefit: zero (0) to high (((((() | Explanation of rating and aspects of the policy option most relevant in this context |

absolute | relative to BAU |

1. Positive impact on SMEs and consumers | ((( | (( | SMEs and consumers are likely to benefit most from facilitation of damages claims under the Preferred Option, especially from introduction of opt-in collective and representative actions • but slight risk that the “loser pays” rule (fee-shifting only recommended) will discourage SMEs and consumers in low-probability cases or, due to the disclosure rule, claimants who do not possess much evidence at the outset.[152] |

2. Likely macro-economic impact | (( | (( | Second pillar of enforcement likely to emerge but less effective than under Policy Option 2 consequently, more limited push for more competitive markets with likely positive effects on growth and employment • very low risk of excessive litigation leading to a deteriorating business environment.[153] |

Summary comparison of Policy Options

193. The likely positive and negative impacts of Policy Options 1 to 5 were set out in the tables in section 5.2, while the impact of the Preferred Policy Option was assessed in section 6.1. The summary set out below provides an overview of the impact, in absolute terms, of the Preferred Policy Option compared with the five other Policy Options.

Table 21: Summary of impacts (in absolute terms) of Policy Options 1-5 and the Preferred Option

Benefits achieved/problem addressed | Rating of benefits from zero (0) to high (((((() |

2. Awareness, deterrence, enforcement and legal certainty | (((( | ((((( | ((( | ( − ((( | ( | (((( |

3. Access to justice | ((((( | ((((( | ((( | ( − ((( | ( | (((( |

4. Efficient use of judicial system | ((( | ((((( | ((( | ( − ((( | ( | (((( |

5. A more level playing field | ((((( | (((( | ((( | 0 − ((( | 0 | (((( |

Costs | Costs from zero (0) to high (((((() |

2. Administrative burden | ((((( | ((( | (( | 0 − (( | 0 | (( |

3. Error costs | ((( | (( | ( | ( | ( | ( |

4. Implementation costs | ((((( | (((( | (( | 0 − (( | 0 | (( |

Other impacts | Rating of other impacts from zero (0) to high (((((() |

2. Macro-economic impact | (( | ((( | (( | 0 − (( | 0 | (( |

Proportionality and EU added value of the Preferred Option

194. Some of the measures envisaged in the Preferred Option entail mere codification of the acquis communautaire , others would alter the status quo . Questions of proportionality and subsidiarity arise only in the case of the latter, i.e. the recommendations for actual changes.

195. The Commission considers that action at EU level along the lines of the Preferred Option would be in line with the principle of subsidiarity since there is a clear need for and added value in such action. The Preferred Option is also fully in line with the principle of proportionality, both as regards its general approach and the content of the individual measures envisaged.

Proportionality

196. As explained earlier, it was particularly because of its qualities in terms of proportionality that the Preferred Option scores higher than alternative Policy Options. In particular, the Preferred Option strikes a careful balance between effective protection of victims’ rights to compensation, the legitimate interests of potential defendants and third parties and important interests of Member States.

197. Above all, the Preferred Option is the minimum necessary to achieve its objective, namely to guarantee that across the EU victims of infringements of EC competition law have access to a truly effective mechanism for obtaining full compensation for the harm they suffered. Of the Policy Options scoring sufficiently well on the objective of effective compensation for a large number of victims, the Preferred Option implies the lowest costs.

198. Moreover, the costs imposed on citizens and businesses are proportionate to the stated objective. To give one example, limiting disclosure of evidence to disclosure by category protects parties from being subjected to frivolous, excessive demands for documents (“fishing expeditions”) and from the spiralling costs often associated with discovery rules in certain other jurisdictions. In particular, access to evidence, as envisaged in the Preferred Option, contains a built-in mechanism for ex-ante judicial control of the necessity for and proportionality of the disclosure measure.

199. Likewise, the costs to Member States linked to the Preferred Option are also proportionate to the objective of effective relief for the victims of antitrust infringements. For each issue in the Preferred Option where there was a choice of possible effective measures, the measures that are least intrusive while still conducive to achieving the main objective were chosen. Attention was paid to avoiding a disproportionate conflict with existing legal traditions, in particular with mandatory rules of ordre public in the Member States (for example, in the context of multiple damages).

European added value

200. The Preferred Option would have European added value for the following reasons[154]:

201. Lack of Community action would conflict with the requirements under Community law, as recalled particularly by the Court of Justice in the Courage and Manfredi judgments, namely to provide for an effective framework for compensation for victims of infringements of Articles 81 and 82. As there is no indication that a significant number of Member States are likely to introduce, in the foreseeable future, legislative changes that ensure an effective legal framework for damages actions by victims of antitrust infringements, only further incentives at European level can create a legal framework ensuring effective redress.

202. There is currently marked inequality in the level of judicial protection of individual rights under the Treaty in the Member States. For example, the absence of Community rules on the issue of collective and representative claims means that consumers and small undertakings in some Member States are at a significant disadvantage over their counterparts in other Member States, in terms of their ability to obtain compensation for infringements of the Community competition rules. Another example is the absence, in many Member States, of clear rules on the issue of the passing-on defence and claims by indirect purchasers. The result is an evident disparity in the very content of the entitlement, under Community law, to damages. More specifically, a claim under the law of one Member State may lead to full recovery of the claimant’s loss, while a claim for an identical infringement in another Member State may lead to a significantly lower award or even no award at all.

203. These sometimes large differences in the level of legal protection of the rights which victims derive from the Community competition rules distort the competitive environment for businesses, as the likelihood and scope of claims for damages against undertakings directly affect their competitive position. Isolated initiatives by Member States are, by nature, not likely to produce a more level playing field for businesses and to reduce the uncertainty created by the current big differences between the national legal systems. On the contrary, individual initiatives may even widen the gaps and increase the risk of a negative impact as a result of forum shopping. Consequently, only Community action can correct the distortions of competition that currently exist due to the variations in the legal arrangements for antitrust damages in the Member States.

204. Application of Articles 81 and 82 of the EC Treaty invariably involves a situation where trade between the Member States is affected by market conduct. Conduct infringing the Community competition rules therefore tends, by nature, to have a transnational dimension that could benefit from a measure at European level. One practical example are the diverging rules on the pass-on defence (e.g. one MS excludes defence and standing for indirect purchasers, whereas another MS does the opposite). Due to the high likelihood of multi-state cases, there is a serious risk that the contradictory rules might lead either to multiple liability for the same overcharge or to no compensation at all. Another example is the rule on the binding effect of NCA decisions across the EU. Such mutual recognition of NCA decisions is clearly a mechanism that is best introduced by a single consistent measure rather than by unilateral action by 27 Member States.

205. The interaction between the measures facilitating antitrust damages actions and various aspects of public enforcement of competition rules needs to be addressed, for instance operation and protection of the Commission’s and Member States’ leniency programmes. Individual action by Member States does not seem capable of achieving this in any consistent manner.

Monitoring and evaluation

206. The research and consultation exercise that led to adoption of the White Paper ranks among the most extensive efforts of this kind. Two extensive studies, public consultations and a series of other consultations with stakeholders at Member State and Community levels, including public authorities, prominent scholars and practitioners from the private sector, have contributed greatly to the analysis and evaluation of the relevant issues to date.

207. Following publication of the White Paper, the Commission will continue and intensify this dialogue with private and institutional stakeholders and with academics and expert practitioners. In the process, the Commission will pay particular attention to monitoring developments, both legislative and judicial, in the legal framework for antitrust damages actions in the Member States.

208. On this basis the Commission will evaluate whether, and, if so, to what extent, to put forward any legislative proposal or other measures such as recommendations or guidance on best practices to enhance the effectiveness of antitrust damages actions.

209. In the first step of the monitoring process, the Commission will be opening a public consultation on the White Paper. Once more, the purpose of this exercise is to obtain the widest possible range of information and opinions in relation to, on the one hand, the measures put forward in the White Paper, be they proposals that would necessitate legislative action or those that entail adoption of recommendations and guidelines, and, on the other, developments in the Member States.

210. The Commission will also actively engage in the continued institutional dialogue with the European Parliament and the European Economic and Social Committee, both of which commented on the Commission’s Green Paper and made recommendations for further action.

211. At this stage, a considerable number of conferences, seminars and other discussions are already scheduled for the period after publication of the White Paper. These will provide the Commission with opportunities to receive feedback from, and exchange views with, representatives from industry, consumer associations, law firms and economic consultants, but also representatives of Member States’ governments, competition authorities and judiciary. When selecting the events in which they participate, the Commission departments will pay particular attention to achieving the widest possible spread in terms of groups of stakeholders and experts and of geographical coverage.

212. The Commission will also use the existing framework for cooperation and discussion with Member States’ competition authorities within the European Competition Network to continue monitoring antitrust damages actions across the EU. Another helpful mechanism in this context is Article 15(2) of Council Regulation 1/2003 which requires Member States to send the Commission copies of all judgments where national courts apply Article 81 or Article 82 of the EC Treaty. It should, however, be borne in mind that a simple count of the numbers of (successful) antitrust damages cases will not provide an accurate measure of the effectiveness of the legal framework for antitrust damages actions. What matters is full and adequate compensation of the greatest possible number of victims of infringements of competition law. Without knowing how many victims have been harmed and how much damage they suffered, the number of successful cases does not tell much. As no data are available on the number of victims, the amount of damage they suffered and the amount of damages paid to victims following judgments or settlements, it is essential that the Commission gathers information by means of an intensive dialogue with a large number of stakeholders.

[1] http://ec.europa.eu/atwork/programmes/docs/clwp2007_en.pdf.

[2] http://ec.europa.eu/comm/competition/antitrust/others/actions_for_damages/study.html. Page numbers quoted in this report refer to the electronic version of the Comparative Study.

[3] See p. 1.

[4] 2006/2207(INI)). The Resolution is available on the website of the European Parliament at: http://www.europarl.europa.eu/oeil/file.jsp?id=5378362.

[5] (INT 306). The Report is available on the website of the EESC at: http://eescopinions.eesc.europa.eu/viewdoc.aspx?doc=\\esppub1\esp_public\ces\int\int306\en\ces1349-2006_ac_en.doc.

[6] Invitations were issued via the Association of European Competition Law Judges, the European Network of Councils for the Judiciary and the Network of the Presidents of the Supreme Judicial Courts of the EU.

[7] Invitation to tender COMP/2006/A3/012 published on 23/12/2006 under reference No 2006/S 245-262369.

[8] SEC(2005) 791.

[9] For an estimate of the harm suffered by consumers, see the Competition Council’s Decision 05-D-65 of 30 November 2005, paragraph 338.

[10] See www.oft.gov.uk/news/press/2007/134-07.

[11] See the estimate by Maarten Pieter Schinkel (University of Amsterdam) reported by NRC Handelsblad at: www.nrc.nl/europa/article773658.ece/Kratje_pils_was_jarenlang_1_euro_te_duur.

[12] This action was brought following the Commission decision of 4.7.2007, Case COMP/38.784 – Wanadoo España/Telefónica.

[13] See http://www.ausbanc.es.

[14] Case C-453/99, Courage and Crehan, [2001] ECR I-6297, and Joined Cases C-295/04 to C-298/04, Manfredi , [2006] ECR I-6619.

[15] Some academic commentators and some representatives of Member States were also more sceptical. For the formal consultation on the Green Paper see the written submissions available at: http://ec.europa.eu/comm/competition/antitrust/actionsdamages/green_paper_comments.html.

[16] See Impact Study, Part III, section 2.1.

[17] European Parliament Resolution (see footnote Error! Bookmark not defined. ), points D and F together with points 11 to 27; Report of the European Economic and Social Committee (see footnote 5), point 5.4.

[18] See Comparative Study, p. 1-1.

[19] See page 1. See also the Resolution of the European Parliament which concludes that there is only “rare and exceptional use of private actions” in competition cases for damages (see point F).

[20] On these obstacles, see also the European Parliament Resolution, points 11 to 27 and Report of the European Economic and Social Committee (see footnote 5), point 5.4.

[21] See also Impact Study, Part I, section 1.

[22] See also Impact Study, Part I, section 2.1.

[23] Case 26/62, van Gend en Loos [1962] ECR (Eng. Spec. Ed.) 1: “(…) the vigilance of individuals to protect their rights amounts to an effective supervision in addition to the supervision entrusted to the diligence of the Commission and of the Member States.”

[24] Case C-453/99, Courage and Crehan (see footnote 14), paragraph 27 and Joined Cases C-295/04 to C-298/04, Manfredi (see footnote 14), paragraph 91.

[25] Case C-126/97, Eco Swiss [1999] ECR I-3055, paragraph 36, and Manfredi (see footnote 14), paragraph 31.

[26] See, for instance, the Commission’s Annual Reports on Competition Policy at: http://ec.europa.eu/ comm/competition/annual_reports/.

[27] See Impact Study, Part III, section 2.1.

[28] Hardcore cartels are outright agreements between competitors to fix prices or allocate markets. The estimates do not cover other infringements of Article 81 such as vertical restraints nor abuses under Article 82.

[29] This estimate is based on the total amount of fines imposed by the European Commission on cartels (annual average for the period 2002 to 2007) and the finding that, on average, the total overcharge applied by these cartels is approximately 50% of the fine. Assuming a given detection rate of cartels (10% on the least conservative assumptions and 20% on the most conservative), the total overcharge applied by undetected EU-wide cartels can be calculated. Another constituent of harm to consumers is the “deadweight loss”. Assuming a set relation to the overcharge applied (50% on the least conservative assumptions and 10% on the most conservative), the total deadweight loss is calculated and then added to the total overcharge in detected and undetected cartels. For further explanations of the method and for data underlying these assumptions plus extensive references to research in this area, see Impact Study, Part I, section 3.1.2.

[30] The deadweight loss mentioned above does not include the surplus lost by the cartelist as a result of the lower quantity sold.

[31] This estimate is based on the figures for (at least) EU-wide cartels (see above) and on the assumption that domestic cartels imply harm to consumers equivalent to 88.4% of the harm resulting from EU-wide cartels; for details on the methodology and underlying assumptions see Impact Study, Part I, section 3.2.1.

[32] See for these estimates and the underlying analysis Impact Study, Part I, section 3.2.1 (Table 10).

[33] See Impact Study, Part I, sections 2.1.1 and 3.1.1.1.

[34] See Impact Study, Part I, section 4.1.

[35] See Comparative Study, pp. 1-72 to 1-75 and 1-110.

[36] The empirical data used by the Impact Study in this comparison are mostly from the USA, a country which has an enhanced system of antitrust damages actions. A range of refinements were, however, made to these data. In particular, the figures mentioned above are not based on treble damages as customary in the USA, but rather single damages with pre-judgment interest, to reflect the predominant legal situation in many Member States (on average, single damages with pre-judgment interest can be said to equate roughly to double damages without pre-judgment interest, see Impact Study, Part I, section 6 with further references).

[37] See previous footnote.

[38] See Impact Study, Part I, section 6.

[39] See the more detailed analysis in the tables in sections 5 and 6.

[40] See Impact Study, Part I, section 6.

[41]9ABPQRpqrs•ø ! % / 0 1 5 < ? A B S T U Ê Ë Í Î |"#$%&'+ùõêãßõÛãêãêãêõÛõÛßÛßÛßõßõÛõÔÐĶ­¶¤™¤‹¤z‹q‹™¤hDmHnHu[pic] [42]?jÓ[pic]ha/VU[pic]mHnHu[pic]jha/VU[pic]mHnHu[pic]ha/VmHnHtH u[pic]ha/VmHnHu[pic]huhWÉhWÉmHnHu[pic]hWÉ

hñVûhñVûhñVûhu

h@>h@>jh@>h@ See Impact Study, Part II, section 1.1 (for a general model of incentives to sue) and Part II, section 3.2 for the specific issue of access to evidence.

[43] See Comparative Study, pp. 1-26 to 1-102.

[44] See Comparative Study, pp. 1-78 to 1-80, 1-111 to 1-112 and 1-127 to 1-129.

[45] See Comparative Study, pp. 1-26 to 1-102.

[46] See Comparative Study, pp. 1-51 to 1-53.

[47] There is a serious risk of a company infringing competition rules being sued for damages in the Member State where the company is incorporated and that not only the procedural rules of that country apply but also the rules of substantive law. See Article 2 of Regulation 44/2001 and Article 6(3)(b) of Regulation 864/2007.

[48] The most important principles in this context (see the Court’s judgment in the Manfredi case (see footnote 14 above)) are equivalence and effectiveness: national rules on the exercise of a Community law right to damages cannot be less favourable than those governing similar domestic actions (principle of equivalence), nor can they render exercise of these Community law rights practically impossible or excessively difficult (principle of effectiveness). For further details see Commission Staff Working Paper, Annex to the Green Paper: Damages actions for breach of the EC antitrust rules, SEC (2005) 1732, points 18-22.

[49] See Joined Cases C-295/04 to C-298/04, Manfredi (see footnote 14), paragraph 95.

[50] See Impact Study, Part II, section 3.2.1 and Comparative Study, pp. 1-53 to 1-76 and 1-107 to 1-111.

[51] See Impact Study, Part II, section 2.

[52] See Comparative Study, pp. 1-43 to 1-47 and 1-106. See also the study on alternative means of consumer redress other than redress through ordinary judicial proceedings, conducted by the University of Leuven, at: http://ec.europa.eu/consumers/redress/reports_studies/comparative_report_en.pdf.

[53] See Comparative Study, pp. 1-70 to 1-71 and 1-110.

[54] See Impact Study, Part II, section 7.

[55] See Comparative Study, pp. 1-88 to 1-91.

[56] See Comparative Study, pp. 1-51 to 1-53 and 1-107.

[57] See Impact Study, Part II, section 4.

[58] See Comparative Study, pp. 1-81 to 1-88 and 1-113 to 1-114.

[59] On the issue of damages, see Comparative Study, pp. 1-93 to 1-97 and 1-116 to 1-118.

[60] See Impact Study, Part I, section 1.

[61] See below.

[62] E.g. companies purchasing claims for compensation and pursuing these claims in their own name before civil courts.

[63] See Impact Study, Part I, section 1.

[64] For example, the 7th amendment to the German Act against Restraints of Competition, in 2005, introduced a range of provisions aimed at improving the conditions for damages actions, particularly those following a decision by the competition authority. Also, the UK Enterprise Act, in force since 2003, extended the jurisdiction of the specialised Competition Appeal Tribunal (CAT) to hear damages claims following a decision of the Office of Fair Trading or the European Commission. It also allows a damages claim to be brought before the CAT by a representative body on behalf, and with the consent, of two or more consumers.

[65] For instance, the Office of Fair Trading recently recommended that the UK Government should, in particular, conduct consultations on whether (and, if so, how best) to allow representative bodies to bring stand-alone and follow-on representative actions for damages or injunctions on behalf of businesses and consumers in competition law. It also recommended that representative actions on behalf of the victims at large should be allowed. In order to protect the success of its leniency programme, the OFT suggested the possibility of removing leniency applications from disclosure to civil parties and removing leniency applicants from joint and several liability. See: “Private actions in competition law: effective redress for consumers and business – Recommendations from the Office of Fair Trading”, available at: http://www.oft.gov.uk/shared_oft/reports/comp_policy/oft916resp.pdf.

[66] See Impact Study, Part II.

[67] Standing to sue means the entitlement given by a legal system to an individual or group to bring legal proceedings before a court.

[68] Joined Cases C-295/04 to C-298/04, Manfredi (see footnote 14), paragraph 61.

[69] See Impact Study, Part II, section 1.6.2.

[70] Joined Cases C-295/04 to C-298/04, Manfredi (see footnote 14), paragraph 93.

[71] The choice of the most appropriate of the possible EC legislative instruments (i.e. the choice between a regulation, a directive or a combination of both) will ultimately depend on (i) the exact content of the Commission proposal, (ii) the legal, economic and political context at the time when the proposal is made and (iii) the degree to which a level playing field in the EU is considered necessary or desirable.

[72] Except for indirect purchasers, see below.

[73] See Joined Cases C-295/04 to C-298/04, Manfredi (see footnote 14), paragraph 97 and Case C-271/91, Marshall [1993] ECR I-4367, paragraph 31.

[74] See Impact Study, Part I, section 6 and Part II, section 4.2.4.

[75] An opt-in collective action is a form of collective redress whereby only those who explicitly agreed to join the collective action are bound by the eventual judgment or settlement. An opt-out collective action, by contrast, is one where all members of a group are bound by an eventual judgment or settlement, unless they explicitly declare otherwise (“opt out”).

[76] “Notice pleading” differs from “fact pleading” in that claimants must give only enough information to suggest that they have a plausible case, as opposed to setting out in detail the facts of their case and describing the relevant means of evidence.

[77] Decisions of the European Commission are already binding under current law (see section 2.3).

[78] See footnote 70.

[79] In this context, the date of conclusion of public proceedings would be the date on which the claimant can reasonably be expected to have knowledge of the decision.

[80] See footnote 70.

[81] The discretionary power of the court extends to both the possibility and the scope of this portion; the court’s decision partially to shift costs can be based on considerations linked to the limited financial resources of the victim, creation of unnecessary procedural costs by one party, the novelty of the legal issue at stake, etc.

[82] As in Policy Option 2, the date of termination of the public proceedings would be the date on which the claimant can reasonably be expected to have knowledge of the decision.

[83] See Joined Cases C-295/04 to C-298/04, Manfredi (see footnote 14), paragraph 95.

[84] See, in particular, Impact Study, Part III, section 2.2 and Part II, sections 1.3, 1.6, 2.3.2, 2.4.2, 5.1.3 and 6.3.1.

[85] See, in particular, Impact Study, Part III, section 2.2, Part II, sections 1.6, 2.3.2, 2.4.2, 3.2 and 5.1.3 and Part I, sections 2.1.1 and 3.1.1.1.

[86] See, in particular, Impact Study, Part III, section 2.2 and Part II, sections 2.3.2 and 3.2.

[87] See, in particular, Impact Study, Part III, section 2.2 and Part II, section 2.3.2.

[88] See, in particular, Impact Study, Part III, section 2.2 and Part II, sections 1.2, 1.6.3, 3.2, 5.1.3 and 6.3.1.

[89] See, in particular, Impact Study, Part III, section 2.2 and Part II, sections 1.6.3, 2.4.2, 3.2.1 and 5.1.3.

[90] See, in particular, Impact Study, Part III, section 2.2 and Part II, sections 1.2, 1.6.3, 2.4.2, 3.2.1 and 5.1.3.

[91] See, in particular, Impact Study, Part III, section 2.2 and Part II, sections 1.6.3, 2.4.2, 3.2.1 and 5.1.3.

[92] See, in particular, Impact Study, Part III, section 2.2.1.

[93] See, in particular, Impact Study, Part III, section 2.2.2.

[94] See, in particular, Impact Study, Part III, section 2.3 and Part II, sections 1.3, 1.6.2, 2.4.1, 2.4.3, 3.2.1, 5.1.1 and 6.3.1.

[95] See, in particular, Impact Study, Part III, section 2.3 and Part II, sections 1.3, 1.6.2, 2.4.1 and 2.4.3.

[96] See, in particular, Impact Study, Part III, section 2.3 and Part II, sections 2.4.1, 2.4.3 and 3.2.1.

[97] See, in particular, Impact Study, Part III, section 2.3 and Part II, sections 2.4.1, 2.4.3 and 3.2.1.

[98] See, in particular, Impact Study, Part III, section 2.3 and Part II, sections 1.6.2, 2.4.1, 2.4.3, 3.2.1 and 6.3.1.

[99] See, in particular, Impact Study, Part III, section 2.3 and Part II, sections 1.6.2 and 3.2.1.

[100] See, in particular, Impact Study, Part III, section 2.3 and Part II, sections 1.2, 1.6.2 and 3.2.1.

[101] See, in particular, Impact Study, Part III, section 2.3 and Part II, sections 1.6.2 and 3.2.1.

[102] See, in particular, Impact Study, Part III, section 2.3.1.

[103] See, in particular, Impact Study, Part III, section 2.3.2.

[104] See, in particular, Impact Study, Part III, section 2.4 and Part II, sections 1.6.1, 2.4.3, 3.2.1, 5.1.1 and 6.3.1.

[105] See, in particular, Impact Study, Part III, section 2.4 and Part II, sections 1.6.1, 2.4.3, 3.2.1 and 4.2.1.

[106] See, in particular, Impact Study, Part III, section 2.4 and Part II, sections 2.4.3 and 3.2.1.

[107] See, in particular, Impact Study, Part III, section 2.4 and Part II, sections 1.2 and 2.4.

[108] See, in particular, Impact Study, Part III, section 2.4 and Part II, sections 1.6.1, 3.2.1 and 6.3.1.

[109] See, in particular, Impact Study, Part III, section 2.4 and Part II, sections 1.6.1 and 3.2.1.

[110] See, in particular, Impact Study, Part III, section 2.4 and Part II, sections 1.2, 1.6.1 and 3.2.1.

[111] See, in particular, Impact Study, Part III, section 2.4 and Part II, sections 1.6.1, 2.4.3 and 3.2.1.

[112] See, in particular, Impact Study, Part III, section 2.4.1.

[113] See, in particular, Impact Study, Part III, section 2.4.2.

[114] See, e.g., Impact Study, Part III, section 2.5.

[115] See, in particular, Impact Study, Part III, section 2.1 and Part II, sections 1.6.1, 2.2.1, 3.2.1 and 4.2.1.

[116] See, in particular, Impact Study, Part III, section 2.1 and Part II, sections 1.6.1 and 4.2.1.

[117] See, in particular, Impact Study, Part III, section 2.1 and Part II, sections 1.6.1, 2.2.1 and 3.2.1. See also the report “Private actions in competition law: effective redress for consumers and business - Recommendations from the Office of Fair Trading”, available at:http://www.oft.gov.uk/shared_oft/reports/comp_policy/ oft916resp.pdf. Regarding the issue of passing-on, the report underlines the need for consistency at EU level. On the issue of the binding force of NCA decisions, it states that a set of unilateral initiatives by the Member States could complicate rather than simplify matters.

[118] See, in particular, Impact Study, Part III, section 2.1 and Part II, sections 2.2.1 and 3.2.1.

[119] See, in particular, Impact Study, Part III, section 2.1 and Part II, sections 1.6.1, 2.2.1 and 4.2.1.

[120] See, in particular, Impact Study, Part III, section 2.1 and Part II, sections 1.6.1 and 4.2.1.

[121] See, in particular, Impact Study, Part III, section 2.1 and Part II, sections 1.6.1 and 4.2.1.

[122] See, in particular, Impact Study, Part III, section 2.1 and Part II, sections 1.6.1 and 4.2.1.

[123] See, in particular, Impact Study, Part III, section 2.1.1.

[124] See, in particular, Impact Study, Part III, section 2.1.2.

[125] See section 4.2.1.

[126] See section 5.2.1.

[127] See Impact Study, Part III, section 2.2.4.

[128] See section 4.2.2.

[129] See section 5.2.2.

[130] These problems between clients and their lawyers arise mainly in the context of settlements, as the agents may have incentives to settle sooner, for lower amounts or under otherwise unfavourable conditions (e.g. certain coupons instead of monetary compensation) than in the interest of the victims (the principals).

[131] See Impact Study, Part III, section 2.3.4.

[132] See section 4.2.3.

[133] See sections 5.2.2 and 5.2.3.

[134] See sections 5.2.1 and 5.2.3.

[135] See Impact Study, Part III, section 2.4.4.

[136] See section 4.2.4.

[137] See section 5.2.4.

[138] See Impact Study, Part III, section 2.5.1 and Table 79.

[139] See section 4.2.5.

[140] See section 5.2.5.

[141] See Impact Study, Part III, section 2.1.

[142] See, for example, Impact Study, Part II, section 1.6.2.

[143] See, in particular, Impact Study, Part II, sections 3.2.1.

[144] See section 4.2.5.

[145] See sections 5.2.2, 4.2.3 and 5.2.4.

[146] See sections 5.2.2 and 5.2.3.

[147] See sections 5.2.2, 5.2.3 and 5.2.4.

[148] See sections 5.2.2, 5.2.3 and 5.2.4.

[149] See sections 5.2.2, 5.2.3 and 5.2.4.

[150] See sections 5.2.2 and 5.2.3.

[151] See sections 5.2.2, 5.2.3 and 5.2.4.

[152] See sections 5.2.2 and 5.2.3.

[153] See 5.2.2, 5.2.3 and 5.2.4.

[154] See sections 5.2.2, 5.2.3 and 5.2.4.

[155] According to point 5 of the Protocol on the application of the principles of subsidiarity and proportionality, “[t]he following guidelines should be used in examining whether the [principle of subsidiarity] is fulfilled:− the issue under consideration has transnational aspects which cannot be satisfactorily regulated by action by Member States;− actions by Member States alone or lack of Community action would conflict with the requirements of the Treaty (such as the need to correct distortion of competition or avoid disguised restrictions on trade or strengthen economic and social cohesion) or would otherwise significantly damage Member States’ interests;− action at Community level would produce clear benefits by reason of its scale or effects compared with action at the level of the Member States” (OJ C 340, 10.11.1997, p. 10).