Alma Mater Studiorum – Università di Bologna - AMS Dottorato

Alma Mater Studiorum – Università di Bologna - AMS Dottorato

Alma Mater Studiorum – Università di Bologna DOTTORATO DI RICERCA IN EUROPEAN DOCTORATE IN LAW AND ECONOMICS Ciclo 27° Settore Concorsuale di afferenz...

1MB Sizes 0 Downloads 15 Views

Recommend Documents

Alma Mater Studiorum Università di Bologna - AMS Dottorato
partire dai 60 anni, il 50-70% dei pazienti con DS sviluppa demenza senile ... di placche senili; il danno ossidativo ed

Alma Mater Studiorum – Università di Bologna - AMS Dottorato
6.7 (http://david.abcc.ncifcrf.gov/) were performed (using the high ..... Prigmore, E., Donahoe, P. K., Smith, R. S., Pa

Alma Mater Studiorum – Università di Bologna - AMS Dottorato
cocoa solids in cocoa butter, while milk chocolate contains milk solids and milk ... than 35 % total dry cocoa solids, i

Alma Mater Studiorum – Università di Bologna - AMS Dottorato
Mar 5, 2010 - 3.3.4 Plasma treatment of electrospun separators . 35. Chapter 4 .... charge/discharge cycles . ...... The

ALMA MATER STUDIORUM - UNIVERSITA' DI BOLOGNA SCUOLA
and the phenomenon of daigou and counterfeit purchase, I analyzed Prada marketing mix strategy in China with product, pr

ALMA MATER STUDIORUM - UNIVERSITÀ di - AMS Tesi di Laurea
love so much in Disney's rendering, was nothing more than the story of an .... is the only tale where the princess alrea

Le Antichità di Bologna di Bartolomeo della Pugliola - AMS Dottorato
Bartolomeo della Pugliola, frate dell'ordine dei Minori Conventuali, è mandato dai superiori del suo .... BARONE, La le

tesi dottorato defitivita - AMS Dottorato
GABRIELE BALDUCCI. Prof. FABIO PEZZI ..... del prodotto (dato dal fatto della grande quantità di prodotto ammostato) e

Schema frontespizio teso Dottorato - AMS Dottorato
List of Figures. 6 .... African Union‟s Inter-african Bureau of Animal Resources ..... forced by destitution to depend

Facsimile del frontespizio della tesi di dottorato - AMS Dottorato
Mar 11, 2015 - of oil is equivalent to ten minutes of operation at low power or three ...... sidering the average per ca

Alma Mater Studiorum – Università di Bologna DOTTORATO DI RICERCA IN EUROPEAN DOCTORATE IN LAW AND ECONOMICS Ciclo 27° Settore Concorsuale di afferenza: 12/F1; 12/E1; 12/E3 Settore Scientifico disciplinare: IUS/15; IUS/14; IUS/13; IUS/05

TITOLO TESI THE APPLICATION OF EUROPEAN COMPETITION LAW IN ARBITRATION PROCEEDINGS

Presentata da: Penio Penev GOSPODINOV

Coordinatore Dottorato

Relatore

PROF. LUIGI ALBERTO FRANZONI

PROF.DR. ROGER J. VAN DEN BERGH

_________________________

_________________________

Esame finale anno 2015

The Application of European Competition Law in Arbitration Proceedings Penio Penev Gospodinov European Doctorate in Law & Economics Erasmus University of Rotterdam A thesis submitted for the degree of Doctor of Philosophy Defence date 15 December 2014

Abstract

This work provides several policy proposals capable to strengthen the private enforcement of EU competition law in arbitration. It focuses on the procedural law aspects that are permeated by legal uncertainty and that have not yet fallen under the scrutiny of the law and economics debate. The policy proposals described herein are based on the functional approach to law and economics and aim to promote a more qualified decision making process by: adjudicators, private parties and lawmakers. The resulting framework of procedural rules would be a cost-effective policy tool that could sustain the European Commission’s effort to guarantee a workable level of competition in the EU internal market. This project aims to answer the following broad research question: which procedural rules can improve the efficiency of antitrust arbitration by decreasing litigation costs for private parties on the one hand, and by increasing private parties’ compliance with competition law on the other hand? Throughout this research project, such broad question has been developed into research sub-questions revolving around several key legal issues. The chosen sub-research questions result from a vacuum in the European enforcement system that leaves several key legal issues in antitrust arbitration unresolved. The legal framework proposed in this research project could prevent such a blurry scenario from impairing the EU private enforcement of competition law in arbitration. Therefore, our attention was triggered by those legal issues whose proposed solutions lead to relevant uncertainties and that are most suitable for a law and economics analysis.

to my wife Irene

Acknowledgements

I would like to gratefully acknowledge the European Union institutions and the European taxpayers for founding, hence making possible, my Ph.D. work in the last three years.

Preface 0.1

Motivation

Competition law is essential for the functioning of the European Union Treaties and for achieving the goals set therein. This statement, embodied in the case-law of the European Court of Justice itself,1 is based on the wording of Article 3 of the Treaty of the European Union (hereinafter Treaty of the European Union (TEU)), namely the very article that states the aims of the European Union, which reads: ‘[The European Union] shall work for the sustainable development of Europe based on balanced economic growth and price stability, a highly competitive social market economy, aiming at full employment and social progress’.2 Article 3 does not only refer to the existence of a positive correlation between economic growth and the level of competition in a given market. If we read it in light of the Freiburg school of economics’ influence on the European Union cultural and economic background, Article 3 implies that competition is the necessary ground upon which social justice can be pursued. In other words, the European Union is founded on the assumption that a sustainable economic growth in Europe passes necessarily through market integration tied to the concept of workable competition. On these premises the European Union produced, on the one hand, a body of substantive law defining a framework of workable competition that is pursued 1

Case C-126/97, Eco Swiss China Time Ltd v Benetton International NV [1999] I3055, available at this URL address: http://eur-lex.europa.eu/LexUriServ/LexUriServ.do?uri= CELEX:61997J0126:EN:HTML. 2 Article 3(3) of the Treaty on European Union, available at this URL address: http:// eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX:12012M/TXT.

iii

0. PREFACE

by the lawmaker, and on the other hand, a body of procedural law determining to what extent the workable competition objective can be, and is, practically achieved. Leaving aside any discussion about the content of the EU competition policy, that is the substantive part, this research is focused on the latter, namely the procedural one. This decision is based on the assumption that, a legal norm is only as good as the mechanism by which it is enforced. Therefore any attempt to strengthen the EU competition law should not overlook the legal enforcement procedure. The traditional antitrust enforcement systems are both public and private. Private enforcement contributes to pursue effectively and efficiently competition law objectives alongside with public enforcement. In the words of Mario Monti as European Commissioner for Competition, private enforcement has the potential, if coordinated with public enforcement, to significantly contribute to an ideal competition law enforcement model which combines both the pillars, private as well as public, and promotes citizens as the principal guardians of the legal integrity of competition law in Europe.3 Nowadays, however, European competition law is mostly enforced by competition authorities and subject to the review of courts. In fact, it is widely acknowledged that private enforcement of both European and national competition law has been extremely limited in Europe.4 Hence, it is possible to argue that the current European procedural law framework does not provide private parties with the incentives that are necessary to achieve the optimal level of private enforcement. Strengthening the European private enforcement has been at the centre of the debate on competition law for almost ten years.5 However, on the one hand, the 3

Monti, Mario. ‘Private Litigation as a Key Complement to Public Enforcement of Competition Rules and the First Conclusions on the Implementation of the New Merger Regulation.’ speech given in Brussels, 2004. 4 See the Study on the conditions of claims for damages in case of infringement of EC competition rules, available at this URL address: http://ec.europa.eu/competition/antitrust/ actionsdamages/study.html. 5 The debate started with the Commission [2005, 2008, 2011] and has recently resulted in the Directive on damages actions for breach of the EU antitrust rules, proposed by the Commission and available at this URL address: http://ec.europa.eu/competition/antitrust/ actionsdamages/documents.html.

iv

lawmaker’s attention and that of the legal scholars focuses mainly on the private enforcement in front of national courts and through mass litigation, leaving almost untouched the topic of private enforcement through any alternative dispute resolution mechanism, such as arbitration.6 On the other hand, the economic literature focuses either on competition policy or on theories of litigation, but it also completely disregards the market players’ incentives when competition law disputes are solved by arbitrators. The need for a Law and Economics analysis on antitrust arbitration is proven by the number of specialised legal practitioners that are investigating the application of competition law in arbitration.7 Their work shows that in order to apply competition law in arbitration proceedings, legal practitioners need to answer a wide range of open questions emerging from the existing legislative vacuum in this matter. In fact, although competition law is an extremely important set of norms subject to mandatory application, the lawmaker does not provide any guidance on how should they be applied by international arbitral tribunals. The adjudicators of different jurisdictions have tried to fill the legislative gap in the field of antitrust arbitration. Nevertheless, the solutions proposed to most problematic issues are often contrasting and inconsistent among jurisdictions. Such scenario increases the legal uncertainty and leaves room for the strategic behaviour of private parties. It creates cracks and loopholes in the structure of this legal field that, due to the fundamental importance of competition law for the EU legal systems, can be as dangerous as cracks and loopholes in a dike. The risk is to jeopardise the level of enforcement of competition law in the European Union and create adverse effects (i.e. distortions) on the whole functioning of the internal market. 6

Although the EU has recently showed interest in alternative dispute resolution mechanisms, for instance introducing the Directive 2013/11/EU of the European Parliament and of the Council of 21 May 2013 on alternative dispute resolution for consumer disputes, available at the URL address: http://eur-lex.europa.eu/LexUriServ/LexUriServ.do?uri=OJ:L:2013:165: 0063:0079:EN:PDF. 7 EU and US Antitrust Arbitration. A Handbook for Practitioners, edited by: Gordon Blanke, Philip Landolt, Kluwer Law International, February 2011.

v

0. PREFACE

Increasingly wary of the contrasting solutions adopted by adjudicators at the national level, legal scholars have tried to fill the legislative gap and to provide with some degree of legal certainty. Comparative law experts have approached this problem examining dispute resolution systems from different jurisdictions, to identify similarities and differences among possible alternative solutions. Comparativists have proposed legal reforms which result from their attempt to understand and to weight advantages and disadvantages of alternative procedural rules. The challenge faced in these studies is that comparative law does not provide with any weighing tool capable to objectively determine the effects of alternative legal rules on the private parties’ behaviour. Comparative law scholars have mostly ignored the literature on the economic analysis of litigation which provides the needed weighing tools. We focus on what is missing in the current status of the legal literature debate, namely, the use of the economic analysis of law. In this research project, our intent is to use the economic analysis of law to study the effects of the contrasting legal solutions proposed by the legal practitioners to overcome the existing legal vacuum in this field. In fact, Law and Economics makes use of a scientific methodology, consisting of mathematically precise theories (e.g. price theory, game theory, cost-benefit analysis) and empirically sound methods (e.g. statistics and econometrics), to verify the effects of different procedural rules on the private parties’ economic incentives. Law and Economics is used here as a methodological tool that can favour a more qualified decision making. The results of this study cast new insights on the effects of the available competing legal rules upon private parties’ behaviour incentives. The implied assumption is that the use of Law and Economics’ methodology for this study can provide a framework of legal solutions complying with the EU legislator’s demand of an effective private litigation process as well as with a higher degree of compliance with the law. Therefore neither lawmakers nor adjudicators should overlook this type of analysis. The few interdisciplinary works on arbitration produced by Law and Economics scholars try either to adapt the litigation theories to arbitration or to

vi

study under which conditions US private parties insert arbitration clauses in their contractual relationships.8 However, these works do not have a European perspective nor specifically address any issue raised by competition law disputes. The primary motivation for this research is to fill the gap in the interdisciplinary debate about the effectiveness of private enforcement of EU competition law through arbitration.

0.2

Methodology

The methodology adopted in this thesis, for the economic analysis of legal rules, rests on what has been called a functional approach to Law and Economics.9 In this comparative evaluation of alternative legal solutions, we focus on both private parties’ economic incentives and on the concept of methodological individualism. The former implies that the analysis starts from the assumption that individuals’ economic incentives can be used to highlight the parties’ preference of one legal rule above the others. Whereas the idea behind the latter relies on the existence of a theoretical market for legal rules and on the fact that, if allowed, rational players will choose the rules that benefit them the most. Therefore, in light of the functional approach to Law and Economics, observing or hypothesising parties’ preferences should allow us to identify failures in the law. Provided that this methodology cannot be profitably applied unless the private parties reveal their preferences within a “market” of legal rules; the field of international commercial arbitration is particularly suited for the use of such approach. In fact, when opting for arbitration, individuals are free to choose both, which alternative legal systems and even which specific legal rules will regulate their business relationship. The same holds true also for antitrust arbitration, even if the EU competition law is one of the few branches of unified substantive EU law, hence somehow impairing the private parties ability to choose between 8 9

Benson [1999]; Drahozal and Wittrock [2008]; Eisenberg and Miller [2006]; Shavell [1995]. Parisi [2004]; Parisi and Luppi [2012]; White [2009].

vii

0. PREFACE

alternative rules. As a matter of fact, EU Member States have different rules on arbitration and they are also allowed to apply competition law under differently fashioned procedural rules. Thus the EU uniformity on the side of substantive law is accompanied by profound differences on the side of procedural law. As we will see below, these differences, reflected in antitrust arbitration, affect the incentives of private parties and force them to reveal their preferences unlacing the full potential of the functional approach to Law and Economics in the ground of antitrust arbitration. Focusing our attention on the procedural law instead of the substantive one provides another methodological advantage from the viewpoint of Law and Economics. The European Union does not refer to perfect competition but instead to the concept of workable competition, as the level of competition to be pursued in the internal market. This is a second best choice; in fact, the theoretical concept of perfect competition is not the optimal level of competition in a given market. The reason is that perfect competition in the internal market is not achievable in practise. One of the constraints is the limited amount of resources available to competition authorities, which makes it impossible to detect and stop all infringements; but, more importantly, it can be proved that there may be cases of efficient breach of competition law. While these restraints are overlooked by the concept of perfect competition they are included by the concept of workable competition. However, while perfect competition can be easily defined, trying to define what is, or should be, the workable level of competition leads towards an unsettled debate revolving around the policy implications of competition law. On the contrary, shifting our attention on the procedural law side, we leave behind the debate on the EU competition policy goals and focus on the goal of optimal enforcement of competition law. As Miller puts it in his 1997 seminal work entitled ’The legaleconomic analysis of comparative civil procedure’,10 when a legal system defines a set of procedural rules that regulate legal enforcement procedures it faces a trade-off between two costs in litigation: ’the costs of the procedure in question and the costs of error’.11 Summary and cheap procedures increase the risk of fact10 11

Miller [1997]. Miller [1997].

viii

finding inaccuracy or erroneous legal analysis. Conversely the cost of these errors can be brought to zero but only at the price of burdensome and costly procedural rules. Therefore an optimal enforcement system can be reached adopting such procedural rules that minimise the sum of these two costs. This also means that cheaper procedural rules that do not raise the cost of legal errors would increase the efficiency of an enforcement system. Moreover, in the private parties’ perspective under a Coasian approach, legal enforcement is not a productive activity but only a transaction cost affecting private parties incentives on their way to obtain what they are entitled to according to the substantive law.12 This implies that a more efficient enforcement system decreases the cost that private parties have to pay in litigation to obtain substantial justice. Thus, the collateral effect of this efficiency increasing cheaper litigation would also be an increase of private parties’ demand for litigation. In other words such an optimal enforcement system could allow to adjudicate worthwhile cases that would not have been otherwise brought to court; hence increasing the deterrence of violations and resulting into a higher degree of compliance with the legal standard set by the substantive law.13 To summarise, the methodology used for this research revolves around the idea of using the functional approach in Law and Economics to determine the optimal procedural rules in antitrust arbitration. The preferred viewpoint is the one of private parties. In fact the functional approach to Law and Economics has the advantage of shifting our attention mainly on the relationship between individual preferences and indirectly on how they affect expected social outcome. The implied assumption, analysed later, is that rational private parties will design an arbitration clause to favour, among possible alternative rules available in arbitration, those that decrease the cost of litigation without increasing the cost of legal errors, facilitating enforcement of competition law. In fact, minimising costs 12

Economists studying litigation theory, adopting the private parties’ viewpoint refer to litigation using the concept of unproductive rent-seeking competition. While adopting the viewpoint of welfare economics one could argue that litigation is a productive activity, since the body of case-law resulting from the litigation activity reduces the uncertainty of a legal system. Nevertheless in light of the methodology used in this research we focus on the private parties’ perspective. 13 As per not so meritorious cases that could also be litigated more often, it will be discussed further.

ix

0. PREFACE

of enforcement for private parties indirectly provides them with incentives leading towards a higher compliance with the law, which corresponds to the preferable social outcome. In line with the European doctrine on procedural law, the aim is not to make parties litigate more, but to make them comply with the established law. As a concluding remark on the methodology we note that the economic and legal theories used in this thesis refer to legal and economic concepts, discussions and arguments that are developed elsewhere in the specialised literature. We will not define them in detail or address their correctness but only refer to the relevant literature for the convenience of the less expert reader. Moreover we will try to tell whether these elements are widely accepted, or not, and how is the state of the debate.

0.3

Problem Definition

In light of the methodology discussed above and given the existing legal gap about antitrust arbitration as described in the motivation section above, our research question can be formulated as follows: When applying EU competition law in arbitration, what are the procedural rules that could improve the efficiency of private enforcement and increase the compliance with competition law? The answer to this broad question clearly depends on which of the many controversial issues of antitrust arbitration we choose to study. Within the introduction chapter this broad research question will be segmented in sub-questions that can describe more specifically the problem definition of this work. In essence, the chosen goal to be pursued in this research project is the reduction of private parties’ transaction costs for enforcing EU competition law in arbitration, while also guaranteeing that the social interests embodied in the competition policy are not overlooked. In fact, enhancing the efficiency of private

x

enforcement will increase the level of deterrence in breaching EU competition law. This goal will be pursued through a comparative evaluation of the alternative procedural rules that can be used to solve key issues in the field. The use of the Law and Economics methodology will guarantee that this research project has a sound theoretical framework and that it could be used for a future empirical legal analysis. Therefore, even if a different goal should be pursued by policy makers, the results of this research will not lose relevance. In fact, this project will highlight the economic effects of different procedural rules on parties’ incentives to breach EU competition law rights or seek their enforcement.

0.4

Limitation of This Research

In this thesis we adopt a European viewpoint. Competition law is one of the few areas in which there is a truly uniform European law, since Member States’ legislation is relegated to a secondary role. However, it is true that arbitration is a matter of procedural law, and as such, although somehow homogeneous, it maintains substantial differences throughout Europe. For this reason we will treat only procedural issues that are of a super-national scope and that are based on cross-country uniformity. The efficiency of substantive rules of competition law is not evaluated in this research. The on-going debate in Europe about competition policy and policy goals that should be pursued through competition law is also outside of the scope of this research. We also disregard the debate of private versus public enforcement, since we undertake the vision that both of them, if coordinated, can independently and simultaneously contribute to the creation of an ideal competition law enforcement system. Other important issues that are not included in this research pertain to the debate on whether arbitration is better than litigation as well as all advantages and disadvantages of arbitration in comparison to litigation. This debate is out of the scope of the present work because these are two different mechanisms designed to satisfy the demand of justice of different market players. For example, whereas ordinary litigation can be in absolute terms cheaper than arbitral proceedings, the opportunity cost associated to the length

xi

0. PREFACE

of court decisions can widely exceed the cost of arbitration procedures for a business party. Therefore, the net result may be in favour of arbitration for economic actors while in favour of court litigation for others. Further research on this topic should be of empirical nature, it should aim to provide more detailed insights into the cost-benefits of enforcing competition law in arbitration. Moreover further research could aim to support a more accurate evaluation of the effects of antitrust arbitration on social welfare, especially in relation to national litigation and public enforcement.

0.5

Social Relevance

There is a massive amount of data as well as economic theories proving that competition law infringements produce negative effects on society and wealth. At the same time, it is undisputed that European private enforcement of competition law is extremely underdeveloped; hence some corrective actions have been called for by a growing number of scholars, practitioners, institutions and market players. Any corrective action taken with the purpose to strengthen the private enforcement of European competition law should aim for the big catch. This research project points in the right direction for two main reasons. Firstly, the recent debate focuses on mass litigation to improve consumers’ protection. However, in the United States where private enforcement proceedings are nine times more than public ones and where class actions are favoured, statistical data proves that, although competition law exists supposedly for the main benefit of consumers, consumers are not the majority of claimants in private enforcement proceedings. One third of the plaintiffs are defendants’ competitors, more than one third are dealers or distributors and less than 20% are consumers.14 This means that the majority of competition lawsuits pertain to contracts between perpetrator and victim of an anticompetitive conduct. This is most likely a result of the 14

White [1985].

xii

fact that every vertical agreement, R&D agreement, price agreement, JV agreement, merger agreement and many others, potentially include an issue of private enforcement competition law of contractual origin. Secondly, in a global economy all the aforementioned business-to-business contractual relationships, even if they are in absolute numbers less than the business-to-consumer ones, are very likely to involve significant amounts of capitals and to concern multinational corporations. In the last 10 to 15 years, alternative dispute resolution mechanisms allegedly gained the title of exclusive jurisdiction for international contractual transactions of the kind mentioned above, i.e. vertical agreements. If two companies do not have their headquarters in the same country, there is good chance that any dispute that may arise upon the performance of a contractual obligation falls within the scope of an arbitration clause. It is worth reminding that the 1958 New York Convention on the Recognition and Enforcement of Foreign Arbitral Awards (hereinafter 1958 New York Convention on the Recognition and Enforcement of Foreign Arbitral Awards (NYC)), ratified by more than 140 countries, is one of the United Nations’ most successful international trade law treaties. In light of the above, the topic of this research can be qualified as aiming at the big catch. It is worth recalling that part of the mission of the European Union is to guarantee a level of workable competition in the European internal market, for the achievement of the goals embodied in the European Treaties. As mentioned above, the achievement of this goal passes necessarily through antitrust enforcement which is subject to the limited amount of resources available to competition authorities. Strengthening private enforcement of competition law through private arbitration is a cost-efficient enforcement mechanism for competition authorities, which is even more valuable in years of economic crisis and increasing budgetary constraints.

0.6

Scientific Relevance

Arbitration seems to have all the characteristics necessary to provide an efficient competition law enforcement mechanism. However, it will still profit from a higher degree of certainty on side of procedural law. Given the flexibility of the

xiii

0. PREFACE

proceedings and the legislative vacuum on this matter, it is all left into the hands of the academic debate, the literature, the practitioners and the case-law. This research brings together several streams of legal and economic literature, including the literature on litigation theory, game theory, transaction costs, cost-benefit analysis, comparative Law and Economics. Through this work, we attempt to overcome the legislative gap in this field and to build an integrated theoretical framework capable to lead towards the optimal level of private enforcement of competition law through arbitration. The ultimate academic intent is to prove why certain legal rule are preferable to other available rules proposed by legal practitioners; and we aim to justify such choice in light of parties’ incentives to comply with competition law. The integrated framework that will be developed in this research will represent a useful tool to study any other areas of law that are considered a matter of public policy and object of extraterritorial application. For instance the enforcement in arbitration of intellectual property rights, bankruptcy rules or other areas of law could face the same issues faced by competition rules. Namely, as it happened to antitrust law, in the future these legal fields may be declared, by the European Court of Justice, ‘of fundamental importance for the European legal system’.15

0.7

Expected or Desired Impact

After the comparative economic analysis of procedural rules applicable to antitrust arbitrations, we will try to suggest a normative framework capable of providing powerful procedural law incentives for the parties. Parties’ incentives due to cost-saving based efficiencies, should increase the effectiveness of the private enforcement of competition law in arbitration. This framework should promote a fully-fledged contribution of arbitration to the private enforcement of competition law. Victims of an anticompetitive conduct could benefit and profit from the significant number of advantages potentially made available by antitrust arbitration. Among the main advantages of this private enforcement mechanism 15

See the Introduction section below for further clarifications this point.

xiv

can be enumerated: timely compensation for loss suffered; punctual civil sanctions of nullity in contractual relationships; competition law skilled adjudicators; compensation of legal costs; administrative discretion avoidance; interim relieves; punitive damages; discovery rules; etc. Consequently, the set of rules which will result as outcome of our research could, for instance, strive for being implemented in any contractual arbitration clause. Such antitrust friendly arbitration clauses can be negotiated by private parties either among themselves or with the European Commission. In the latter case the European Commission could promote such clause both when it decides a case under Article 9 of Regulation 1/2003 or when it clears a merger case with behavioural commitments. The framework of legal rules emerging from this research project could also be implemented into the arbitration rules of private and public international arbitration institutions. Provided that the latter aim to guarantee an economically more effective enforcement mechanism for conflicting parties in a vertical agreement. Nevertheless it could also be a reference point for the alignment of contrasting orientations by national courts or international arbitral tribunals. In conclusion, since the policy proposals emerging from this research strive to be an effective way to strengthen private enforcement rules at low political/budgetary costs, the European lawmaker’s future regulatory initiatives could be inspired and take into consideration the results emerging from this work.

xv

0. PREFACE

xvi

Contents Preface

iii

0.1

Motivation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

iii

0.2

Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

vii

0.3

Problem Definition . . . . . . . . . . . . . . . . . . . . . . . . . .

x

0.4

Limitation of This Research . . . . . . . . . . . . . . . . . . . . .

xi

0.5

Social Relevance . . . . . . . . . . . . . . . . . . . . . . . . . . . .

xii

0.6

Scientific Relevance . . . . . . . . . . . . . . . . . . . . . . . . . . xiii

0.7

Expected or Desired Impact . . . . . . . . . . . . . . . . . . . . . xiv

Contents

xvii

List of Figures

xxiii

List of Cases

xxv

Glossary

xxix

1 Introduction 1.1

1

EU Competition Law . . . . . . . . . . . . . . . . . . . . . . . . .

2

1.1.1

The Central Rules of the European Antitrust Policy . . . .

2

1.1.1.1

Article 101 TFEU . . . . . . . . . . . . . . . . .

3

1.1.1.2

Vertical Restraints . . . . . . . . . . . . . . . . .

3

1.1.1.3

Article 102 TFEU . . . . . . . . . . . . . . . . .

5

1.1.1.4

Concentrations and Merger Control . . . . . . . .

5

Enforcement of Competition Law . . . . . . . . . . . . . .

7

1.1.2.1

7

1.1.2

Public and Private Enforcement . . . . . . . . . .

xvii

CONTENTS

1.2

1.1.2.2

The Foundation of the EU Private Enforcement .

10

1.1.2.3

Private Enforcement in EU National Courts . . .

14

1.1.2.4

Private Enforcement in Arbitration . . . . . . . .

17

International Arbitration . . . . . . . . . . . . . . . . . . . . . . .

19

1.2.1

The Arbitration Proceedings . . . . . . . . . . . . . . . . .

21

1.2.1.1

The Arbitral Tribunal . . . . . . . . . . . . . . .

22

1.2.1.2

Jurisdiction and Arbitrability . . . . . . . . . . .

23

1.2.1.3

Choice of Law . . . . . . . . . . . . . . . . . . . .

25

1.2.1.4

The Arbitration Hearings . . . . . . . . . . . . .

27

1.2.1.5

The Arbitral Award . . . . . . . . . . . . . . . .

28

The Second-look Review of an Arbitral Award . . . . . . .

29

1.2.2.1

Annulment of the Award . . . . . . . . . . . . . .

30

1.2.2.2

The Recognition and Enforcement of the Award .

31

Antitrust Arbitration . . . . . . . . . . . . . . . . . . . . . . . . .

32

1.3.1

Jurisdiction . . . . . . . . . . . . . . . . . . . . . . . . . .

33

1.3.2

Merits . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

34

1.2.2

1.3

2 Arbitral Tribunals’ Jurisdictional Issues in the Application of EU Competition Law 37 2.1

The Legal Debate on Antitrust Arbitration Clauses . . . . . . . .

40

2.1.1

40

The Objective Arbitrability of Competition Law . . . . . . 2.1.1.1

The American Safety Doctrine Denies Antitrust Arbitrability . . . . . . . . . . . . . . . . . . . .

41

2.1.1.2

The Breakthrough Mitsubishi Decision . . . . . .

43

2.1.1.3

The Objective Arbitrability of Competition Law in Europe . . . . . . . . . . . . . . . . . . . . . .

47

The Conflict of Law on the Antitrust Objective Arbitrability . . . . . . . . . . . . . . . . . . . .

51

The Subjective Arbitrability of Antitrust Disputes . . . . .

58

2.1.2.1

The Anticompetitive Arbitration Clause . . . . .

60

2.1.2.2

Is Private Enforcement Within the Scope of a General Arbitration Clause? . . . . . . . . . . . .

66

Antitrust Arbitration Needs a Uniform Choice of Law Rule . . . .

72

2.1.1.4 2.1.2

2.2

xviii

CONTENTS

2.2.1

2.2.2

Conflict of Law and Extraterritorial Application of Competition Law . . . . . . . . . . . . . . . . . . . . . . . . . .

72

2.2.1.1

The Conflict of Law Issue in Antitrust Arbitration 73

2.2.1.2

The Extraterritorial Application of Competition Law . . . . . . . . . . . . . . . . . . . . . . . . .

74

Choice of Law Rules . . . . . . . . . . . . . . . . . . . . .

75

2.2.2.1

National Courts’ Approach to Conflict of Law Issues 76

2.2.2.2

Arbitral Tribunals’ Approach to Conflict of Law Issues . . . . . . . . . . . . . . . . . . . . . . . .

2.2.2.3

78

List of Choice of Law Rules in Antitrust Arbitration 79

2.2.3

Conflict of Law Notions Relevant for Antitrust Arbitration

83

2.2.4

Analysis of the Effect of the Second-look Review on the Suitability of Arbitral Tribunals’ Choice of Law . . . . . .

86

2.2.4.1

Second-look Review Doctrine in the US . . . . .

87

2.2.4.2

Limited and Extended Second-Look Review in Europe . . . . . . . . . . . . . . . . . . . . . . . . .

88

New Challenges to the EU Conflict of Law Approach to Antitrust Arbitration . . . . . . . . . . . . . . . . . . . . .

90

2.2.5

2.2.5.1 2.2.5.2

A Hypothetical Russian Example of Public Policy Exception . . . . . . . . . . . . . . . . . . . . . .

91

Overview of the Russian Approach to Antitrust Arbitration . . . . . . . . . . . . . . . . . . . . .

92

3 The Jurisdictional Issues of Antitrust Arbitration Under the Lens of Game Theory 97 3.1

Correcting a Less Desirable Outcome Through Competition Law Private Enforcement . . . . . . . . . . . . . . . . . . . . . . . . . 100

3.2

An Argument Against Limitations to Antitrust Arbitrability . . . 114

3.3

Should Competition Law Opt-in or Opt-out of an Arbitration Clause?123

3.4

Not all Arbitration Clauses are Alike . . . . . . . . . . . . . . . . 125

3.5

A Possible Solution to the Conflict of Law Issues in Antitrust Arbitration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 136

xix

CONTENTS

4 Arbitral Tribunals’ Powers to Apply EU Competition Law 4.1

4.2

Cooperation Between Arbitral Tribunals and the European Commission . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149 4.1.1

Regulation 1/2003 and the Cooperation Mechanisms With National Courts . . . . . . . . . . . . . . . . . . . . . . . . 150

4.1.2

Arbitral Tribunals and the Cooperation Mechanism of Regulation 1/2003 . . . . . . . . . . . . . . . . . . . . . . . . . 152

Arbitral Tribunals’ Jurisdiction to Request a Preliminary Ruling Under Article 267 TFEU . . . . . . . . . . . . . . . . . . . . . . . 155 4.2.1

The Functioning of Article 267 TFEU . . . . . . . . . . . . 156

4.2.2

Limitations to the Recognition and Enforcement of an Arbitral Award . . . . . . . . . . . . . . . . . . . . . . . . . . 160

4.2.3

The Importance of Preliminary Ruling for Arbitration, When “Public Policy” is at Stake . . . . . . . . . . . . . . . . . . 161

4.2.4

Participation Constraint Analysis . . . . . . . . . . . . . . 162

4.2.5

4.2.4.1

Description of the Structure of the Problem . . . 163

4.2.4.2

General Assumptions . . . . . . . . . . . . . . . . 164

4.2.4.3

Definition of the participation constraint . . . . . 165

Dissipation of Parties’ Effort Analysis . . . . . . . . . . . . 170

5 Arbitral Tribunal’s Duties to Apply EU Competition Law 5.1

5.2

147

181

Legal Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182 5.1.1

The Background of the Eco Swiss Case . . . . . . . . . . . 182

5.1.2

The Principle of Judicial Passivity vs. Antitrust Rules as Public Policy . . . . . . . . . . . . . . . . . . . . . . . . . 184

5.1.3

How the Adjudicator Addressed the Eco Swiss Case . . . . 185

5.1.4

Flaws in the Eco Swiss Case-law and the Legal Practitioners’ Solutions . . . . . . . . . . . . . . . . . . . . . . . . . 188 5.1.4.1

National Courts’ Duty to Apply Antitrust Rules Ex Officio . . . . . . . . . . . . . . . . . . . . . . 189

5.1.4.2

Arbitral tribunals’ Duty to Apply Antitrust Rules Ex Officio . . . . . . . . . . . . . . . . . . . . . . 190

Economic Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . 194

xx

CONTENTS

5.2.1 5.2.2

Background Hypothesis . . . . . . . . . . . . . . . . . . . The Model . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.2.2.1 The Standard Model of Litigation . . . . . . . . . 5.2.2.2 The Model With Arbitration . . . . . . . . . . . 5.2.2.3 When Also Arbitral Tribunals Apply Antitrust Rules Ex Officio . . . . . . . . . . . . . . . . . . 5.2.2.4 When Arbitral Tribunals Refuse to Apply Antitrust Rules Ex Officio . . . . . . . . . . . . . .

194 198 199 202 204 206

6 Conclusions 6.1 The Relevance of The Proposed Legal Framework Resulting from Our Law and Economics Analysis . . . . . . . . . . . . . . . . . . 6.2 The Arbitrability of Competition Law Disputes . . . . . . . . . . 6.3 Pro-competitive and Anticompetitive Arbitration Clauses . . . . . 6.4 Cooperation Between Competition Authorities and Arbitral Tribunals in Conflict of Law Scenarios . . . . . . . . . . . . . . . . . 6.5 Limiting the Second-look Review of an Arbitral Award through the CJEU’s Preliminary Rulings in Arbitration . . . . . . . . . . 6.6 Arbitrators’ Duty to Apply Antitrust Rules of Their Own Motion, a Misfire? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

209

Appendix 1

223

Appendix 2

227

References

231

xxi

209 210 212 215 217 219

CONTENTS

xxii

List of Figures 4.1 4.2

Probabilities attached to all possible outcomes of a CJEU preliminary ruling request in arbitration . . . . . . . . . . . . . . . . . . 166 Rent-Seeking Comparison of Alternative Interpretations of Article 267 TFEU . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 176

xxiii

LIST OF FIGURES

xxiv

List of Cases European Union • Case C-26/62, Van Gend en Loos v. Nederlandse Administratie der Belastingen [1963] ECR 1 • Case 6/64, Flaminio Costa v. ENEL, [1964] ECR 585 ´ • Case C-56/64 and C-58/64, Etablissements Consten S.´a.R.L. and GrundigVerkaufs-GmbH v Commission of the European Economic Community [1966] ECR 299 • Case 48/69, Imperial Chemical Industries Ltd v. Commission (Dyestuffs) [1972] ECR 619 • Case 127/73, BRT v. Sabam, [1974] ECR 51, par 16 and 17 and C-282/95 P Guerin Automibiles v. Commission [1997] ECR 1-1503 • Case C-31/80, L’Oreal v. De Nieuwe AMCK [1980] ECR 3775 • Case C-102/81, Nordsee Deutsche Hochseefischerei GmbH v Reederei Mond Hochseefischerei Nordstern AG & Co KG [1982] ECR 1095 • Case 89/85, Ahlstrm Osakeyhti v Commission (Wood Pulp) [1988] ECR 5193 • Case C-393/92 Gemeente Almelo et al. v. NV Energiebedrijf IJsselmij [1994] ECR I-1277

xxv

0. LIST OF CASES

• Cases C-430/93 and C-431/93 Jeroen van Schijndel and Johannes Nicolaas Cornelis van Veen v. Stichting Pensioenfonds voor Fysiotherapeuten [1995] ECR I-4705 • Case T-102/96 Gencor Ltd v Commission of the European Communities [1999] ECR II 00753 • Case C-126/97, Eco Swiss China Time Ltd v Benetton International NV [1999] I-3055 • Case C-453/99, Courage v. Bernard Crehan [2001] ECR I-6297 • Cases C-295/04 to C-298/04, Vincenzo Manfredi v Lloyd Adriatico Assicurazioni [2006] ECR I-6619 • Case C-8/08, T-Mobile Netherlands BV, and Others v Raad van Bestuur van de Nederlandse Mededingingsautoriteit [2009] ECR I-4529

United States • Prima Pain Corp. V. Flood and Conklin Mfg. Co., 388 U.S. 395 (1967) • American Safety Equipment Corp v. J.P. Maguire and Co., 391 F.2nd 821, 825 (2d Cir. 1968) • The Bremen v. Zapata Off Shore Co., 407 U.S. 1 (1972) and Scherk v. Albero-Culver, 417 U.S. 506 (1974) • Mitsubishi Motors v. Solar Chrysler-Plymouth, 473 U.S. 614, 105 S. Ct. 3346 (1985) • Simula v. Autoliv. 175 F. 3d 716 (9th cir. 1999) • Louis Dreyfus Negoce S.A. v. Blystad Shipping and Trading Inc., 252 F3d 218, 224 (2d Cir. 2001) • Green Tree Financial Corp. v. Bazzle, 539 U.S. 444, 452 (2003)

xxvi

• JLM Industries, Inc. v. Stolt-Nielsen SA, 387 F3d 163 (2d Cir. 2004) • Buckeye Check Cashing, Inc. v. Cardegna, 546 U.S. 440 (2006) • Granite Rock v. Intern. Broth. of Teamsters, 130 S. Ct. 2847 (2010) • Rent-A-Center, W., Inc. v. Jackson, 130 S. Ct. 2772 (2010) • Stolt-Nielsen S.A. v. AnimalFeeds International Corp., 130 S. Ct. 1758 (2010) • ATandT Mobility LLC v. Concepcion, 131 S. Ct. 1740 (2011)

Others • Belgian Supreme Court, 28 Jun 1979 Audi-NSU v. S.A. Adelin Petit, 1 Pas 1260 • Garden Cottage Foods v. Milk Marketing Board [1984] 1 AC 130 (UKHL) • Court of Appeal of Paris in Almira Films vs. Pierrer (1989) • Court of Appeal of Paris, 29 March 1991, Ganz v. Socit Nationale des Chemin de Fers Tunisiens • Court of Appeal of Paris, 19 May 1993, Labinal v Mors • Court of Appeal of Paris, 14 October 1993, Aplix v. Velcro • Cassation Civile, 20 Decembre 1993, Comit populaire de la minicipalit de Khoms El Megreb c. Dalico Contractors, JDI (Clunet) 121 [1994] • Dutch Supreme Court (Hoge Raad) 21 Mar. 1997, NJ 1998, 207 • Court of Appeal of Paris, 18 November 2004, Thal´es AirDefence v. GIE Euromissile (2005) • Court of Appeal of Milan, Terrarmata Tensacciai, 2006

xxvii

0. LIST OF CASES

• Cassation Civile, 7 June 2006 Comproprit maritime Jules Verne v. ABS., JDI (Clunet) [2006] • Premium Nafta Products Limited et al. v. Fili Shipping Company Limited [2007] UKHL 40 • Russian Supreme Commercial Court case A27-4626/2009 OJSC Efirnoe v. LLC Delta Vilmar CIS • Federal Arbitrazh (Commercial) Court of the north-west circumscription case decided on 10 may 2011, case n. A56-68936/2010

Arbitral Awards • ICC Award no. 1397 of 1966 • ICC Award no. 2811 of 1979 • ICC Award no. 7097 of 1993 • ICC Award no. 7673 of 1993 • ICC Award no. 7539 of 1995

xxviii

Glossary ADR Alternative dispute resolutions. 17 CJEU Court of Justice of the European Union. 10 Commission European Commission. 4 ICAC International Commercial Arbitration Court at the Chamber of Commerce and Industry of the Russian Federation. 53 ICC International Chamber of Commerce. 51 Modernisation Regulation Council Regulation 1/2003 of 16 December 2002 on the implementation of the rules on competition laid down in Articles 81 and 82 of the EC Treaty, OJ L 1, 4.1.2003. 9 NYC 1958 New York Convention on the Recognition and Enforcement of Foreign Arbitral Awards. xiv TEU Treaty of the European Union. iv TFEU Treaty of Functioning of the European Union. 3

xxix

Glossary

xxx

Chapter 1 Introduction EU competition law as well as international arbitration are legal fields that have been in-depth analysed by the legal literature.1 This brief overview aims to help the reader who is not familiar with either one of these two topics. This Chapter will firstly introduce the competition law features that will be used and referred to throughout the rest of this work. More specifically, the first Section will be divided in two parts. The first part explains why competition law is so important for our every-day business life and describes what legal rights are embodied in the EU competition rules. The other part defines which mechanisms are used to enforce these competition law rights to prevent and deter a breach of EU competition law. The second Section of this Chapter will focus on international arbitration. In the first part of this Section, the reader will be introduced to the fundamental concepts of this field of law, throughout an overview of the entire proceeding of a theoretical arbitration case. The second part of this Section introduces the procedure of review, recognition and enforcement of an arbitral award in front of national courts.

1

In general about EU competition law see: Roth et al. [2013], Korah [2004], Korah and O’Sullivan [2002], Faull and Nikpay [2014], Jones and Sufrin [2011], Whish and Bailey [2012], Pedro et al. [2011]; whereas in general about international arbitration see Blackaby et al. [2009].

1

1. INTRODUCTION

The Chapter’s final Section is the liaison of the two topics: competition law and international arbitration. When these two fields of law touch and encounter each other, the result is a set of legal issues representing the framework upon which builds this research project. Therefore, the final Section of the Chapter contains an overview of the structure of this work.

1.1

EU Competition Law

The fact that competition law has a history which reaches back to the Roman Empire gives a hint of the relevance that competition law has for the progress of society.2 Starting from the most general definition of competition law, ‘it is a field of law that promotes or maintains market competition by regulating anticompetitive conduct by companies’.3 The lawmaker and the case-law have to clarify in details the meaning of the concepts of market, competition, anticompetitive conduct and companies. Filling these general concepts with a specific content is a necessary act for the practical application of competition rules. The interpretation activity performed by these subjects is substantially defined by the competition policy pursued within a specific legal order; and the goals that a competition authority aims to achieve is what shapes its competition policy. In light of the above, an overview of these goals shows the function of competition law and its importance in the present-day world.

1.1.1

The Central Rules of the European Antitrust Policy

In Europe, there is a multitude of goals that have been pursued over time through the EU competition policy, for instance: market integration (necessary to achieve the four fundamental freedoms embodied in the EU Treaties);4 , protection of consumers’ welfare while guaranteeing also production efficiencies;5 as well as 2

Wilberforce et al. [1966], p. 20. Taylor [2006], p. 1. 4 Free movement of goods, persons, services and capitals, within the EU. 5 This goal was pursued through supporting efficient cost-saving practises along the supply chain. 3

2

the protection of individual economic freedoms.6 Van den Bergh and Camesasca have duly noted that, although today the dominant view on what is the goal of EU competition law revolves around ensuring an efficient allocation of resources and enhancing consumer welfare, some of the current competition rules can be justified only in the light of a different viewpoint.7 In other words, presently, also non-economic goals, such as social equity, play a role in the trade-off of contrasting goals shaping the essence of the EU competition policy.8 From this brief list of the goals that have been and that are still pursued through competition rules, clearly emerges the reason why competition law is so important for our every-day business life. What follows is an overview of the legal protection available against anticompetitive conducts; it aims to clarify how our everyday business activities are, in practise, affected by competition law. Article 101 and Article 102 of the Treaty of Functioning of the European Union (hereinafter Treaty of Functioning of the European Union (TFEU)) along with the Merger Control Regulation are at the cornerstone of the European competition legislation. 1.1.1.1

Article 101 TFEU

Article 101 TFEU defends market players against both horizontal and vertical restraints. Horizontal restraints can derive from agreements, decisions or practises by undertakings that restrict competition between actual or potential competitors who operate at the same level of the supply chain. Most notably, members of a cartel aim to increase products’ prices, fix market shares and impede the marketaccess to new competitors. However, as we will see later, for the purpose of this research it is necessary to focus on vertical restraints. 1.1.1.2

Vertical Restraints

Competition restraints that affect vertical agreements are called vertical restraints. This term can be associated to wide range of business contracts between firms 6

Namely protection of small businesses from being overcome by dominant firms. Van den Bergh and Camesasca [2006], p. 5. 8 Van den Bergh and Camesasca [2006], p. 5. 7

3

1. INTRODUCTION

operating at different levels of the production or distribution chain. A nonexhaustive list of these contracts includes: exclusive and selective distribution agreements, franchising, agency, supply agreements, and many others. The European Commission (hereinafter also referred to as the European Commission (Commission)) has emphasised the relevance of a competition law breach in this kind of contracts because almost all products reach the final customers through a distribution channel formed by vertical agreements.9 In fact business practise is pervaded with examples of vertical agreements. Sooner or later, any company is likely to ask (or to be asked by a counterparty) for the inclusion in such vertical agreements of certain competition restrictions (i.e. vertical restrictions).10 Vertical restrictions can assume various forms and few examples would be: price fixing (in which the manufacturer sets a specific retail price for a product), territorial protection (limiting the places where each dealer may sell the goods), quantity fixing (which sets a level of minimum or maximum units that can be sold), customer restrictions (defining the group of buyers to whom goods can be sold).11 In the past, overlooking the importance of economic theories, competition authorities considered vertical restraints dangerous for welfare, competition and consumers. Nowadays, embracing the more recent economic studies, it is widely accepted that restrictions in vertical agreements can be justified in light of allocative efficiencies; hence firms are allowed to use vertical restrictions, under certain circumstances.12 Those clauses, in fact, can solve coordination problems in vertical structures and limit the scope of opportunistic behaviours by firms. However, it has been proved also that ‘they may have ambiguous effects on economic welfare, depending on the context in which they are used.’13 Whether a specific vertical restraint is anticompetitive, or not, depends upon the competition policy pursued by a specific decision maker. For instance, many vertical restraints which will be considered innocuous in the US are forbidden in the EU in light of the 9

Commission [2002], p. 7. and also Korah and O’Sullivan [2002], Part 1, p. 28. Utton [2005], pp. 233-272. 11 Rey and Caballero-Sanz [1996], pp. 3-10. 12 Rey and Caballero-Sanz [1996], pp. 11-21. 13 Van den Bergh and Camesasca [2006], p. 206; although it is out of the scope of this work, in Van den Bergh and Camesasca [2006], p. 206 ff. is possible to find a general explanation of the Law and Economics debate on vertical restraints. 10

4

EU internal market’s need to contrast territorial segmentation.14 1.1.1.3

Article 102 TFEU

Whereas Article 101 TFEU, described above, provides protection against anticompetitive practises perpetrated by two or more independent firms, Article 102 TFEU prohibits some types of unilateral behaviour by firms holding a dominant position on a determined market. This provision targets conducts considered to be an abuse of a dominant position. Article 102 TFEU is relevant for this research because anticompetitive practises can be perpetrated also through vertical restraints, when dealing with firms that operate at different levels of the production or distribution chain. For instance, through the so-called exclusionary behaviour a dominant firm commonly aims to harm actual or potential competitors operating at the same level of the supply chain by obstructing their access to the upstream or downstream market. This result can be achieved through specific contractual clauses imposed to the upstream or downstream weaker party in vertical agreements. Examples of this type of vertical anticompetitive contractual clause include price discrimination, rebates, tying and bundling, refusals to deal and predatory pricing. However, what has been said above on vertical restraints under Article 101 TFEU, can be valid also for these clauses. They can generate efficiencies and benefit consumers, hence deserving to be allowed under certain circumstances. In practise, it is not easy to assess whether there are, or there are not, positive welfare effects. Such evaluation, both, affects and reflects the goals that a competition policy aims to achieve. What is important to keep in mind is that Article 102 TFEU, the second main rule of EU competition law, protects the internal market from anticompetitive vertical restraints by a firm abusing its dominant position. 1.1.1.4

Concentrations and Merger Control

The third, and last, main branch of EU competition law is called merger control. Unlike the two provisions discussed above, which are included in the Treaty of Functioning of the European Union, this subject matter is addressed by the 14

Zekos [2008b].

5

1. INTRODUCTION

Regulation 139/2004.15 As we have seen, usually companies interact with business partners, even on a long-term basis, through vertical agreements. No matter how stable such relationship can be, in any case, those agreements are not final. Therefore, when there is a breach of competition law, the agreement can be cancelled and its effects revoked. Conversely, under certain circumstances, vertical agreements may not be suitable to satisfy the needs for the natural evolution and growth of a firm. Since, long-term contracts can be cancelled, the business activity may require a more permanent integration of two companies’ assets and know-how. Such permanent integration is possible through mergers (two merging companies create a new company), acquisitions (a company purchases all the assets of another company) or joint ventures (when two companies pool part of their assets into a new entity). All those activities are defined in the EU competition law as concentrations. On the one hand, a concentration can have positive effects, expanding markets and bring benefits to the economy: new products can be developed or produced more efficiently; production or distribution costs can be reduced; the companies can increase their competitiveness in the market, resulting in a higher-quality goods at better prices for consumers. On the other hand, a concentration may also reduce competition strengthening the merging companies’ power in a market. The anticompetitive effects that can be caused by a concentration are usually divided in two categories: unilateral effects and coordinated effects. The former can be the effect of the reduced number of market players. The two merging companies are not competitors after the concentration and they can jointly exploit their market power. For instance, they can profitably rise the price of one or both of their products harming consumers.16 The latter can be described as the effect of the improved changes and possibilities for the inferior number of market players to coordinate their behaviour producing anticompetitive effects. In both these cases, unlike what happens with vertical agreements, the reduction of competition, due 15

Council Regulation 2004/139/EC of 20 January 2004 on the control of concentrations between undertakings (the EC Merger Regulation), OJ L 24/1 16 Ivaldi et al. [2003], Werden [2003].

6

to a concentration, has permanent effects. It is extremely difficult to reverse a concentration ex post. Merger control Regulation 139/2004 provides the necessary legal basis to prevent and guarantee ex ante, before a concentration occurs, the risks of negative effects on competition due to a concentration.

1.1.2

Enforcement of Competition Law

The previous Section briefly overviews the reason why competition law is so important for a modern society and how it can affect market players’ everyday business lives. It also introduces the main parts of EU competition law and describes the three main types of anticompetitive conducts contrasted by Articles 101, 102 of the TFEU and by the Merger Regulation. This Section of the introduction deals with mechanisms used to prevent and deter the breach of competition law. There are two instruments which have been used to stop a breach of antitrust law.17 One is public enforcement and the other is private enforcement.18 1.1.2.1

Public and Private Enforcement

Public enforcement means that there is a competition authority which is entitled to investigate, detect, prosecute, and punish any violation of competition law. This means that any natural or legal person found responsible for a breach of competition law can be fined by the authority. This fine has a deterrence function and a quasi-compensation one, due to the harm caused to the state economy. Any individual can inform the competition authority of an alleged or suspected breach of competition law. However, even if the violation and the harm are proven, the competition authority will fine the perpetrator but will not grant damages in favour of the victims of the anticompetitive conduct. Therefore the record fines of hundreds of millions, that we read about in the newspapers’ headlines, benefit the state finances but do not compensate the direct victims 17

These two instruments have been introduced for the first time in the US by the Sherman Act in 1890 (Sherman Antitrust Act, Ch. 647, 26 Stat. 209, codified at 15 U.S.C. §§1-7; Clayton Antitrust Act of 1914, Pub. L. No. 63-212, 38 Stat. 730, enacted October 14, 1914, codified at 15 U.S.C. §1227, 29 U.S.C. §5253.) 18 Jones [1999], p. 14.

7

1. INTRODUCTION

of the breach. In most EU Member States, the nature of these fines is that of an administrative law punishment. On the other hand, the private enforcement of competition law is a litigation procedure in which private parties claim rights based on competition law provisions. Commonly, the civil consequences of a violation of competition law are: • an injunction for a party to cease violating the other party’s rights; • to award of damages; and • to invalidate a contract.

Hence, any person who has suffered damages caused by an anticompetitive conduct has the right to sue the offender demanding monetary compensation and bringing the infringement to an end. Since a legal system cannot tolerate any undertaking in breach of a mandatory legal provision, any agreement that infringes competition law is null, void and, therefore, cannot be enforced. As a result, also this procedure of enforcement punishes the actor of the unlawful conduct and deters him, and others, from future transgressions. In practise, private parties can use their competition law based rights as a shield or as a sword.19 The former occurs, typically, in contractual liability cases. The plaintiff claims the specific performance of a contractual clause and alleges that there was a breach of the contract by the defendant, who relies upon a competition law violation to invalidate the specific clause or the whole contract. Conversely, in the latter case, when used as a sword, the plaintiff directly wants to stop an anticompetitive conduct and seeks compensation for the harm caused. Therefore private enforcement does not only contribute to prevent violations of competition law, alongside with public enforcement, if coordinated with it, private enforcement could significantly contribute to an ideal antitrust enforcement model which combines both the pillars to achieve an optimal enforcement mech19

Valentine Korah, ‘In introduction guide to EC competition law and practise’.

8

anism.20 The citizens could become the guardians of the EU internal market and legal order’s integrity; especially in the context of vertical restraints, due to their privileged viewpoint on harmful effects.21 In the US, both means of enforcement are well developed. As a matter of fact, the private enforcement cases are nine times more than the public enforcement ones.22 Whereas, in Europe, traditionally, public enforcement has been the main instrument used to guarantee compliance with competition law. In the EU legal order, both the European Commission and the National Competition Authorities (hereinafter NCAs) have the function and responsibility to prevent violations of the EU competition law. In respect to private enforcement, some EU Member States have included in their national antitrust legislations this enforcement mechanism. Only recently, the EU legal order itself has, gradually, opened its doors to the private actions for damages based on competition law. First, in 2001, with the ground-breaking Court of Justice decision known as the Courage case.23 Then Regulation 1/2003 was promulgated, the so-called Council Regulation 1/2003 of 16 December 2002 on the implementation of the rules on competition laid down in Articles 81 and 82 of the EC Treaty, OJ L 1, 4.1.2003 (Modernisation Regulation).24 Most importantly with the Directive of the European Parliament and of the Council on certain rules governing actions for damages under national law for infringements of the competition law provisions of the Member States and of the European Union.25 20

Speech by Mario Monti when European Commissioner for Competition matters: Private litigation as a key complement to public enforcement of competition rules and the first conclusions on the implementation of the new Merger Regulation, at the IBA 8th Annual Competition Conference in Fiesole (Italy), on 17 September 2004. 21 Van den Bergh and Camesasca [2006], p. 332. 22 European Commission, ‘The EU gets new competition powers for the 21st century’ (2004) competition policy newsletter - special edition 1. 23 Case C-453/99, Courage v. Bernard Crehan [2001] ECR I-6297, see also the opinion of Mr Advocate General Mischo delivered on 22 March 2001. 24 Council Regulation 1/2003 of 16 December 2002 on the implementation of the rules on competition laid down in Articles 81 and 82 of the EC Treaty, OJ L 1, 4.1.2003. 25 As of today it has been approved by the European Parliament and soon to be approved also by the Council, all informations available at this url:http://ec.europa.eu/competition/

9

1. INTRODUCTION

At this point, it is worth opening a digression about the evolution and foundation of private actions for damages in the European legal system. 1.1.2.2

The Foundation of the EU Private Enforcement

The EU Treaties do not explicitly provide individuals with the right to start private actions to seek damages based on a breach of competition law. Initially, it was not even clear whether the EU competition law provided any rights at all directly to individuals. The existence of a EU law based individual right to damages was declared by the case-law of the Court of Justice of the European Union (hereinafter also referred to simply as the Court of Justice or as the Court of Justice of the European Union (CJEU)). This interpretation of the Treaties is based on several key EU law principles highlighted by the CJEU in its early case-law. Starting form the Van Gend en Loos seminal decision, the CJEU stated that the EC Treaty in its preamble explicitly stated that it referred not only to governments but also to the people of EuropeCase C-26/62, Van Gend en Loos v. Nederlandse Administratie der Belastingen [1963] ECR 1. Regardless of the legislation of a Member State, the EU law is intended to confer upon individuals specific rights which become part of their legal heritage.26 When a EU law provision is qualified as having direct effect, the rights granted therein have direct application in national law, and national courts must protect these rights.27 Finally, the CJEU held that the prohibitions of Articles 101 and 102 TFEU (at that time called Articles 85 and 86 of the Treaty of Rome) tend, by their very nature, to produce this direct effect in relations between individuals.28 Therefore these provisions create direct rights for the individuals concerned, rights which national courts must safeguard. Moreantitrust/actionsdamages/documents.html. 26 Case C-26/62, Van Gend en Loos v. Nederlandse Administratie der Belastingen [1963] ECR 1. 27 Case C-26/62, Van Gend en Loos v. Nederlandse Administratie der Belastingen [1963] ECR 1. 28 Case 127/73, BRT v. Sabam, [1974] ECR 51, par 16 and 17 and C-282/95 P Guerin Automibiles v. Commission [1997] ECR 1-1503.

10

over the CJEU also clarified that these EU based rights are so important that they even prevail upon national legislation. In fact, when there are national laws which have the effect of limiting or preventing such private actions, the principle of supremacy of the EU law comes into play. This principle held by the CJEU in the Costa v. ENEL case states that the EU Treaties have created their own legal system. This legal system ‘became an integral part of the legal system of the Member States, that have permanently limited their sovereign rights and thus, have created a body of law which binds both their nationals and themselves. The law stemming from the Treaty could not, because of its special and original nature, be overridden by domestic legal provisions, however framed, without the legal basis of the community itself being called into question’.29 The possibility of private enforcement in the EC competition law has been first explicitly pronounced by the CJEU in the Courage v. Crehan case.30 This case was saluted as the major achievement in the direction of effective protection of rights conferred on individuals by EU competition law.31 In the Courage v. Crehan case, the Court of Justice does not even consider as an impediment for compensation the fact that the party suing for damages was actually an undertaking of the contract responsible of restricting or distorting competition. In fact, the CJEU states that a contractual party can rely on the breach of competition law before a national court even to obtain relief from the other contracting party. In other words, just agreeing to an anticompetitive contractual clause that favours the counterparty, does not imply waiving the rights of damages caused by it. Furthermore, in this decision the Court of Justice continues stating that ‘The full effectiveness of article 85 of the Treaty (nowadays called Article 101 TFEU) and, in particular, the practical effect of the prohibition laid down in article 85(1) (now Article 101(1) TFEU) would be put at risk if it were not open to any individual to claim damages for loss caused to him by conduct liable to restrict or distort competition’.32 In practice, this enunciates the existence of a EU right 29

Case 6/64, Flaminio Costa v. ENEL, [1964] ECR 585. Case C-453/99, Courage v. Bernard Crehan [2001] ECR I-6297, see also the opinion of Mr Advocate General Mischo delivered on 22 March 2001. 31 Andreangeli [2004]. 32 Case C-453/99, Courage v. Bernard Crehan [2001] ECR I-6297, par 26. 30

11

1. INTRODUCTION

to damages and, by implication, of a principle of civil liability of individuals for a breach of the EU law. This apparent breakthrough in the EU legislation just reflects a common sense general legal principle known to civilisation since the Roman empire. Everyone is bound to make good of loss or damage arising as a result of his conduct in breach of a legal duty, in latin neminem laedere principle.33 To summarise the significance of the Courage v. Crehan case, one could say that it was the starting point for the private enforcement of European competition law. Therein the CJEU, gradually building on the ground of the general principles embodied in its case-law, recognised that the EU Treaties’ competition law provisions confer certain rights to individuals which can be enforced in front of national courts. Eventually, private enforcement became formally part of the EU legislation with the adoption of the Regulation 1/2003 on implementation of the rules of competition laid down in Articles 101 and 102 TFEU. This regulation is the so-called Modernisation Regulation, which created a fully-fledged private enforcement system next to the public one. The Regulation 1/2003 is responsible for the decentralisation of the EU antitrust law enforcement.34 Namely, the power to fully apply Article 101 and 102 is not exclusively in the hands of the EU Commission anymore. As a matter of fact, before Regulation 1/2003, the application of the third paragraph of Article 101 was a monopoly of the EU Commission. Even after the Courage case, national judges theoretically entitled to grant damages could not assess whether a conduct was permitted by the third paragraph of Article 101 TFEU. According to Article 101(3) TFEU (which resembles a EU equivalent of the US rule of reason) conducts that are formally in breach of the first paragraph of Article 101 TFEU, can be exempted from the prohibition set by the first paragraph. In order for a conduct in breach of Article 101(1) TFEU to qualify for the exemption, under 33

Eilmansberger [2004]. Council Regulation 1/2003 of 16 December 2002 on the implementation of the rules on competition laid down in Articles 81 and 82 of the EC Treaty, OJ L 1, 4.1.2003. 34

12

the third paragraph, it has to pass a competition policy trade-off test. The outcome of this test should exempt conducts that are formally anticompetitive but at the same time can substantially contribute to achieve the goals set by the EU competition policy. At the early stage of creation of the European Union, it was believed that the Commission monopoly on this provision was justified in light of the strong link between Article 101(3) and the EU competition policy. Moreover at that time the internal market, the case law and the culture of competition law were not uniformly developed throughout the European Union. In due time those arguments lost their grounds. Now, Article 6 of the Regulation 1/2003 states that national courts shall have the power to apply Articles 101 and 102 (in their entirety). Therefore, it finally suppressed the exemption monopoly held by the EU Commission. Among the reasons for this overturn was the intent to stimulate private parties to have more frequent recourse to national courts in actions for damages.35 Apparently, the EU lawmaker proved to understand the importance of a modern and developed private enforcement mechanism. The national courts were assigned a new role with respect to the EU competition policy. It has been argued that this new decentralised application system was bound to lead to a ‘new rights-based common culture of competition in the Community (now Union), to create a real culture of diffuse competition law enforcement’.36 Moreover, ‘it consolidates the interpretation of the third paragraph of article 81 (now Article 101 TFEU) as a true rule of law and not as a discretionary political tool in the hands of the Commission’.37 In parallel, Regulation 1/2003 lead to a certain “privatisation” of competition policy enforcement. Previously, the EU Commission had to be notified an anticompetitive agreement and decide ex ante whether to exempt it or not. Nowadays, the ex ante notification procedure has been cancelled. Therefore, while national courts assess ex post, after a dispute arises, whether an anticompetitive agreement deserves to be exempted, companies and their legal 35

Wils [2002], pp. 150-154. Wils [2002], pp. 151. 37 Wils [2002], pp. 150. 36

13

1. INTRODUCTION

advisers hold the burden and the risk, to fulfil the competition rules, engaging into a rigorous self-assessment on the matter.38 In practise, the vast majority of cases can be solved using the EU Commission case-law and guidelines, satisfying the need of certainty in the legal system. However under Regulation 1/2003, when there are controversial cases, they will be decided by national courts taking a position with respect to the EU competition policy. To conclude this overview of the establishment of private enforcement in Europe, it is appropriate to quote recital 7 of the Regulation 1/2003: ‘National courts have an essential part to play in applying the Community competition rules. When deciding disputes between private individuals, they protect the subjective rights under Community law, for example by awarding damages to the victims of infringements. The role of the national courts here complements that of the competition authorities of the Member States.’39 1.1.2.3

Private Enforcement in EU National Courts

This Chapter started describing why competition law is important, its content and the rights that it grants to individuals. Then it focused on explaining what are the mechanisms that are used to enforce competition law and their historical evolution in the EU legal system. Until here we introduced the CJEU’s case-law and the Regulation 1/2003 enhancing national courts’ powers to enforce the EU competition law. This Section clarifies that these initial, necessary and positive steps towards a modern, developed and effective, private enforcement system in Europe are not sufficient. Notwithstanding this evolution the private enforcement in Europe remains underdeveloped. In 2004, the European Commission sought a better understanding of the existing obstacles to successful actions for damages in the EU Member States. The international law firm Ashurst carried an extensive study on this matter. The study provides an analysis of the legal framework and the national law conditions 38

Craig and De B´ urca [2011], pp. 1005-1008. Recital 7 of the Council Regulation 1/2003 of 16 December 2002 on the implementation of the rules on competition laid down in Articles 81 and 82 of the EC Treaty, OJ L 1, 4.1.2003. 39

14

required to claim antitrust damages in most of the current European Member States. It describes the private enforcement across EU Member States as an extremely underdeveloped legal tool. Moreover, the study highlights the profound inconsistencies existing among Member States’ jurisdictions in their approach to this legal field.40 Although, a detailed description of the competition law private enforcement in front of Member States’ jurisdictions is outside of the scope of this work, it is worth providing a general background for readers not familiar with this topic. The heart of the private antitrust enforcement in Europe lies on the relationship between European Union and national law. At the current stage of the European integration, EU law based rights and obligations are in principle enforced before national courts under provisions of national procedural law.41 The Modernisation Regulation, did not result in a harmonisation of procedures, national sanctions nor remedies. However according to Professor van Gerven, there are four EU substantive remedies existing at the cornerstone of the EU private enforcement: a general one, to set aside a national measures that conflict with EU law (in light of the EU law general principles of supremacy, direct effect, effectiveness and equivalence)42 and three specific ones, interim relief, restitution and compensation (for damages).43 Starting with the latter, in order to obtain damages, a private party shall go through the following steps: 1. identify the general or specific substantive and procedural statutory basis for his/her private action claim; 40

Waelbroeck et al. [2004], Ashurst Study, dated 31 August 2004. available at this url:http: //ec.europa.eu/competition/antitrust/actionsdamages/study.html 41 Eilmansberger [2004]; and Wils [2005], pp. 46-47. 42 Craig and De B´ urca [2011], p. 256. As mentioned above, under the principle of supremacy (sometimes referred to as primacy) of EU law, the laws of the EU Member States that conflict with laws of the Union shall be ignored by national courts. According to the principle of equivalence, requirements to successfully claim a right based on breach of EU law should not be stricter than those for infringement of national law. The principle of effectiveness states that national procedures must allow full protection of individual rights based on directly applicable provisions of the EU Treaties. 43 Van Gerven [2000].

15

1. INTRODUCTION

2. identify, under the applicable statutory rules, which is the court competent to hear his/her action for a violation of the EU competition law; 3. prove that he/she satisfies the requirements for the standing of natural or legal person in that specific jurisdiction (including territorial jurisdiction); 4. prove a faulty breach of EU competition law; proving the existence of a certain, specific and quantifiable harm and the causation (i.e. the causal link between the harm and the antitrust violation); and 5. oppose any defence raised against the claim, e.g. contributory fault, force majeure, passing-on defence. Although these steps are common to all EU Member States, the so-called Ashurst study highlighted that the requirements underneath each one of these rules can be very different among Member States jurisdictions. For instance, the main inconsistencies in the damages remedy can be of two kinds: (i) how the nature and the measure of damages is determined; or (ii) how fault, causation, and standing shall be proved (i.e. rules of evidence) in order to be entitled to damages. Unfortunately such inconsistencies do not refer only to the damages remedy, the same holds true also for interim reliefs and restitutions. As a result, both the substantive and the procedural conditions for civil antitrust enforcement can be very different among Member States depending on their national rules. In fact, the Ashurst study qualifies inconsistencies and inadequacies in national laws on remedies and procedures, as a source of serious concern for the effectiveness of the competition law private enforcement. Indeed, those differences tend to create variations in enforcement costs of the EU antitrust rules, and thus to produce unequal conditions of competition among EU Member States.44 Some of the solutions proposed to facilitate private enforcement of EU competition law include: • removing limitations to standing; • improving legal certainty, through recourse to expert courts for dealing with difficult competition law matters, as well as increase of cooperation 44

Van Gerven [2000].

16

with competition authority (acting as amicus curiae); • facilitating rules of evidence, provided that it is difficult to prove the various elements of liability; lowering the standard of proof or even a reversal of the burden of proof could prove crucial; and • reducing the costs, caused also by the length of civil action proceedings. The European Commission tried to face some of the challenges mentioned above, publishing first the Green paper and then the White paper on damages actions for breach of the EC antitrust rules.45 These proposals have been followed firstly by a discussion focused on mass litigation private enforcement to improve consumers’ protection and lately also by a proposed directive on private actions for damages. At the present day, the latter has been approved only by the European Council but not by the European Parliament, hence it has not yet entered into force. Nevertheless, most likely, this directive will soon be part of the EU hard law provisions guaranteeing a minimum core of harmonisation on this matter. In any case, the proposed directive overlooks the needs and peculiarities of the business-to-business private actions. 1.1.2.4

Private Enforcement in Arbitration

It is widely believed that instead of national court litigation, business parties may prefer an alternative mechanisms to enforce contracts or recover damages. In fact litigation in front of national judges is not the only way, neither it is the most effective one, to seek legal protection against a breach of law. The legal practise suggests that Alternative Dispute Resolution (hereinafter Alternative dispute resolutions (ADR)) is a suitable tool to solve disputes based on international commercial contracts. This is especially true with long-term ongoing relationships, such as vertical agreements, where specific investments have been made by both 45

European Commission Green paper - damages actions for breach of the EC antitrust rules SEC(2005) 1732, December 2005. URL http://eurlex.europa.eu/LexUriServ/LexUriServ.do?uri=CELEX:52005DC0672:EN:NOT. and European Commission White paper on damages actions for breach of the EC antitrust rules SEC(2008) 404 SEC(2008) 405 SEC(2008) 406, April 2008. URL http://eurlex.europa.eu/LexUriServ/LexUriServ.do?uri=CELEX:52008DC0165:EN:NOT.

17

1. INTRODUCTION

parties of the contract. Arbitration has a prominent role among the means of Alternative Dispute Resolution. Arbitration is a mechanism for resolving disputes outside of a state’s court. In arbitration, the conflicting parties refer the decision of their dispute to an arbitral tribunal composed by one or more arbitrators. Those are private individuals who are requested to make a decision, called arbitral award, which is legally binding for both parties and widely enforceable.46 In other words, an arbitration procedure means that conflicting parties agree to refer the adjudication of their dispute to a third private individual (arbitrator) who acts as a “judge”. When business parties choose arbitration instead of traditional litigation they are looking for an Alternative Dispute Resolution mechanism that satisfies the peculiar needs of the business world. Examples of the key elements of arbitration that make it particularly suitable and attractive for international commercial transactions include: • more flexible, simple and less formal procedures; • a private judge who is both expert in the legal field of the dispute as well as sensible to the business practises; and • faster and certain private solution to a dispute, which reduces all the costs connected to lengthy litigation (i.e. opportunity costs). All those characteristics of arbitration have been recently acknowledged by the EU public bodies. On 12 March 2013 the European Parliament approved the EU Commission’s proposal of two new legislative acts: (i) the proposal for a Directive on Alternative Dispute Resolution for consumers disputes, (ii) the proposal for a Regulation on online dispute resolution for consumers disputes.47 However these pieces of legislation do not focus on business-to-business transactions, since they will be applicable only to disputes between companies and consumers. The rest of 46

The New York Convention on the Recognition and Enforcement of Foreign Arbitral Awards is one of the most successful international conventions, it has been signed and is applied in more than 140 United Nations Member States. 47 See EU Press Release Memo/13/192, at http : //europa.eu/rapid/press−releaseM EM O− 13 − 192e n.htm.

18

this Chapter focuses on a description of the current state of private enforcement through international arbitration.

1.2

International Arbitration

This Section of the Chapter introduces the main features and the general functioning of arbitration proceedings. The next (and last) Section of this Chapter deals with whether and how arbitration can practically be used for increasing the private enforcement of EU competition law. Arbitration has a contractual nature, hence it is based on the parties’ will to opt-out from the state administered litigation system in favour of a private dispute resolution technique. In theory, an arbitration agreement may occur both before or after a dispute arises. In the former case, it is easier for the parties of a commercial transaction to negotiate and agree on the terms upon which to solve a prospective future dispute based on their business relationship. This may not be the case when the two parties are not business partners or more generally when they are not predetermined subjects. As it is often said in the legal practice, one cannot negotiate today with people who might run one down in an automobile tomorrow. Furthermore, it is also difficult to reach an agreement to arbitrate after a conflict arises. After a clear dispute is defined, the two parties are likely to have contrasting interests on the best way to solve the dispute. The positive characteristics of arbitration would have negative effects on the counterparty’s interests. Therefore, in practise, ex ante contractual arbitration clauses can be considered as the most common form of arbitration agreements. After the business parties have negotiated and agreed on the main terms of a contract, before signing it they include an arbitration clause at the bottom. The arbitration clause is considered part of the so-called “mid-night clauses”. The term mid-night clauses is due to the fact that they are the last one to be hurriedly negotiated, since they are not considered so essential for the success of the deal. However, the very day after the contract is signed, the parties can come to regret such “rushed” negotiation on the arbitration clause. In fact, an

19

1. INTRODUCTION

arbitration clause implies many choices from the contracting parties. In some cases the chosen options can deeply influence the outcome of the dispute. For example a standard arbitration clause could read as follows: ‘All disputes arising out of or in connection with the present contract shall be finally settled under the rules of arbitration of [the name of an arbitration institution], by one or more arbitrators appointed in accordance with the said rules.’ Through such a standard arbitration clause, usually not longer than a couple of sentences, the parties of a contract agree upon two main elements of an international arbitration: (i) the extent of the arbitrators jurisdiction, and (ii) the rules which shall govern the arbitration. In order to describe how those two elements can affect the outcome and the effectiveness of the arbitration we need to disentangle the temporal structure of a standard dispute resolution through arbitration. Solving a dispute through international arbitration is a two-phase procedure, and this Section of the Chapter will be divided accordingly. This research project refers to the first phase using the term arbitral proceeding. It includes everything that happens from the moment when a dispute arises until the moment that a solution to the dispute is provided by an arbitral award. Each one of the steps involved will be discussed in light of the parties’ choice in the arbitration clause and their consequences for the dispute resolution. In a temporal order, after a disputes arises, an arbitral procedure is likely to progress according to the following main steps: 1. an arbitral tribunal must be established; 2. the arbitral tribunal shall determine whether it has jurisdiction over the dispute at stake; 3. it must determine the law applicable to the case; 4. it must hear the parties and analyse the evidence; 5. it must pronounce an arbitral award within a defined timeline; Since arbitral tribunals are composed by private subjects, the arbitrators, their final decisions unlike the national courts’ ones, do not have directly the

20

power to command the compliance of private parties. Eventually, without the participation of a state judge the arbitration proceedings remain into the realm of private parties. Hence when a losing party is unwilling to comply, an arbitral award needs to be recognised and enforced by a jurisdiction that has the so-called ius imperium. Otherwise the arbitral award would be just a worthless piece of paper that does not provide a real solution to the dispute. This second phase of an international arbitration in front of a national court is called in this work award second-look review. It includes the necessary steps that need to be taken to have a national jurisdiction exercise its ius imperium against or in support of an arbitral award. The term award second-look review has been chosen because it describes the nature of this second phase procedure. It is the arbitral award that is reviewed and not the dispute at stake. A national court wants to verify that each step of the first phase called arbitral proceedings complies with minimum standards and requirements before recognising/enforcing the award. The arbitral award second-look review can be performed within two procedures:

1. the losing party may request the annulment of the arbitral award; and 2. the winning party may request a recognition/enforcement of the award.

In light of the title of this research, the application of EU competition law in arbitration proceedings, the analysis will focus on the first phase mentioned above. Nevertheless it is necessary to study it in relationship with the other phase. The arbitration proceedings and the award second-look review are two separate proceedings in front of two different adjudicative bodies. However both of them are necessary elements to solve a dispute between conflicting parties.

1.2.1

The Arbitration Proceedings

Each step of the two phases that form an international arbitration will be overviewed below, starting from the arbitration proceedings phase.

21

1. INTRODUCTION

1.2.1.1

The Arbitral Tribunal

When a party decides to trigger the arbitration clause the first step is to notify to the counterparty its intention to start the arbitration procedure. This notice48 functions also as a request for cooperation in appointing the arbitral tribunal. The contractual arbitration clause has to determine the rules to be followed to appoint the arbitral tribunal. There are many options available. In an ad hoc arbitration, the two parties may be required to agree upon the names of the arbitrators, otherwise an independent appointing authority will be in charge of it. The rules to be followed for appointing the arbitral tribunal can be either the rules of arbitration chosen by the parties in the arbitration clause or a set of default national arbitration rules applicable to the dispute. Conversely in case of an administered arbitration, the whole arbitration procedure will be administered by the professional Arbitration Institution initially chosen by the parties in their contractual arbitration clause.49 Those professional Arbitration Institutions have a ready-to-use set of procedural arbitration rules describing not only how the arbitral tribunal shall be appointed, often it will be appointed by the Arbitration Institution, but also how the whole arbitration procedure shall be conducted.50 Typically, arbitrators are legal professionals. The key element for choosing them is their specific “experience” in deciding related arbitration cases, as well as their skills or understanding of that specific legal and business field. As a result, in theory, parties should not be exposed to the same risk as they are when using ordinary courts. National judges called upon to decide a complex EU competition law issues may even have no knowledge nor experience in the field of law that they are required to apply. 48

which in administered arbitration is called Application for Arbitration and is directed to the arbitration institution. For further details see below. 49 Almost every state has an Arbitration Institution, the most famous are the LCIA in London, the ICC in Paris and the American Arbitration Association in the US. 50 In theory the party could blindly opt-in those rules or decide either to modify or to opt-out of some of them.

22

1.2.1.2

Jurisdiction and Arbitrability

The claimant, commencing the arbitration proceedings, attaches to the Application for Arbitration also the Statement of Claim. This document retraces the facts and the evidence upon which the dispute is grounded in the eyes of the claimant. The counterparty will, then, respond to this with a Statement of Defence presenting the facts and the evidence in support of the defendant’s position. Both these deeds will be the starting point from where the arbitral tribunal will conduct the procedure and solve the dispute. At this point, it is necessary to take a step back. In order for the dispute to be validly decided by the arbitral tribunal, the dispute firstly needs to satisfy two conditions: • the dispute must fall within the scope of the arbitration agreement of the parties; and • the dispute must be wholly, or at least partially, capable of being solved by arbitration. Only if both these conditions are satisfied, it can be said that a specific dispute is arbitrable, namely, it can validly constitute the subject matter of an arbitration procedure. The justification for imposing these two conditions lies on the principle of party autonomy and its limits. It is because of the principle of party autonomy that private parties are entitled to opt-out of the state’s administrated litigation system in favour of a private Alternative Dispute Resolution mechanism. The rationale behind the first condition is to verify that the parties’ original intent was to refer to arbitration the very same kind of dispute that concretely arose between them. As a matter of fact, instead of the standard arbitration clause indicated above, an arbitration agreement can be drafted in such a way to include or to exclude from its scope certain types of disputes. For example, an arbitration clause could say that disputes based on the application of EU law are to be excluded from arbitration, hence they have to be decided by national courts.

23

1. INTRODUCTION

Regarding the second condition of arbitrability, it is due to the specific limits set by a legal system to the party autonomy. In some legal systems there is a more liberal approach to subject matters that can be solved in arbitration, in others a more restrictive one. At one extreme we find provisions such as Article 177 of the Swiss Private International Law saying that ‘any dispute of patrimonial character (meaning that has pecuniary value for the parties) can be referred to arbitration’.51 On the other extreme there are legal systems, for instance China, that specifically list main categories of disputes which are excluded from arbitration.52 The justification for those limits relies on the idea that certain kinds of disputes can affect the public interest. Thus, certain rights are based on mandatory rules which are not arbitrable. In other words, a legal system may consider the rights involved in some kind of disputes to be so important that their application should be exclusive jurisdiction of the national judges. The problem with this line of reasoning is that it may lead to absurd results. For instance, in such scenario, an arbitral tribunal cannot acknowledge promptly that a contract in clear violation of a state mandatory rule is null and void. Instead, the arbitral tribunal has to declare that it has no jurisdiction over the dispute and has to refer the parties to national court that is probably overloaded with work and cannot promptly decide the case. Therefore, nowadays the principle that rules of public policy are not arbitrable and that arbitrators cannot apply mandatory laws is extremely limited. The principle of arbitrability has expanded and came to cover even securities law, intellectual property, bankruptcy law, fraud or corruption issues as well as embargo regulations. Examples of disputes still excluded from international arbitration include conflicts based on employment agreements or consumer contracts. The reason for such restrictions is the intent to protect the weaker party’s right to access to justice. Furthermore, it is worth noting that even in more conservative legal systems, arbitrability must be explicitly excluded by a legal provision in order to claim the lack of arbitrators’ jurisdiction. The arbitrability is not affected by general legal 51

Swiss Private International Law Act (PILA), Article 177(1). According to Articles 2, 3 and 77 the Chinese Arbitration Law of 1995, in China disputes with the public administration, disputes over personal rights and labour disputes are not arbitrable. 52

24

provisions establishing the exclusive jurisdiction of a specific national courts over a subject matter. In fact, the purpose of such norms is to define the allocation of powers or division of functions within the state administered judicial system. Exclusive jurisdiction rules are not intended to determine the non-arbitrability, they are norms of procedural nature directed to national courts.53 To sum up, private parties have to reach an ex ante agreement about which of their potential future disputes are going to be settled in arbitration but under some legal systems they may not be free to arbitrate all kind of legal disputes.

1.2.1.3

Choice of Law

At this point the reader may be wondering, firstly, who has jurisdiction to determine whether the two conditions above are met and, secondly, what are the criteria adopted to reach such decision. The former question is easily answered: issues of arbitrability can be brought before a national court or before an arbitral tribunal. The most widely accepted principle of kompetenz-kompetenz allows the arbitral tribunal to determine whether it has jurisdiction over an issue. According to this principle, arbitrators do not have to wait for a court to determine their jurisdiction before or after they decide a dispute. Regarding the latter question about the criteria to be applied to assess the issue of arbitrability, it depends on the law that the arbitral tribunal or the national court deems applicable to the dispute and it will be discussed below. In order to decide a dispute, the adjudicator, no matter whether a judge or an arbitrator, needs to use a set of both substantive and procedural rules to reach an appreciable decision. For instance, in a country belonging to the European continental civil law tradition, if the parties have not agreed otherwise, a national judge will use the civil code and the procedural code or any other default rules applicable to the case in light of a conflict of law rule. However there is also a viable alternative way to determine the law applicable, based on the general 53

Racine [1997], p. 29.

25

1. INTRODUCTION

principle of party autonomy, mentioned above. Parties are allowed to choose among a multitude of both substantive and procedural rules, including a choice of law clause in their contract. On the one hand, both in litigation and in arbitration, the parties can agree that the adjudicator should not decide the case according to the national civil code but instead in light of an alternative set of substantive rules such as the principles of fairness, equity, or the rules of a specific legal system. This is normally called a choice of law clause. On the other hand, unlike traditional litigation, when the parties include a choice of law provision in the arbitration clause, they are also free to choose a set of procedural rules that should govern the arbitration proceedings. These rules will affect the arbitration procedure with respect to: the timeline, the fact finding, the hearings and the guarantees of the right of fair trial. Every arbitration institution provides a framework of arbitration rules which is nothing else than a set of procedural rules applicable to the arbitration. Legal systems also provide national rules on arbitration proceedings that private parties could choose from, which are often inspired by the United Nation’s UNCITRAL Model Law on International Commercial Arbitration.54 As a matter of fact, in arbitration the choice of law goes even a step further. In arbitration the parties can also choose the seat of the arbitration proceedings. As it will be clarified below, this element of an arbitration clause can be very important for the outcome of the dispute. Although in principle arbitration is detached from any specific national legal system, the arbitration proceedings must be conducted somewhere in a physical place, its seat. When the parties agree upon the place where the arbitration will be conducted they are also indirectly agreeing to subject the award second-look review to a national court of that specific legal system. In other words, the seat of the arbitration determines the forum that could be responsible for the annulment of an arbitral award. In light of the above, one could state that in arbitration the parties enjoy an extensive freedom to choose the law applicable to determine the solution of 54

Hoellering [1986].

26

their dispute. Nevertheless, also this perspective of parties autonomy is subject to different limitations across jurisdictions. These limitations are due to conflicts in substantial or procedural law that are not considered acceptable by some jurisdictions. For example during the contract negotiations, the party with the stronger bargaining power could, in theory, force the counterparty to agree upon certain procedural rules that are in breach of the fundamental right of fair trial (for instance we could think about procedural rules that do not provide to both parties the same right of witness cross-examination). The principle of party autonomy cannot go as far as breaching the principle of fair trial. All legal systems assume that there are certain rules of public policy that cannot be disregarded, they are of mandatory application irrespective of parties choice of law. When the arbitral tribunal or the national court have to decide the law applicable to the dispute, they are facing these limitations to the party autonomy. Therefore international arbitrators have to choose the law applicable to the dispute trying to fulfil the parties will and expectations but also considering carefully the solution to potential conflicts of law determined by the application of mandatory rules.

1.2.1.4

The Arbitration Hearings

After the arbitral tribunal has been established, after it has determined its jurisdiction over the case and it has identified the law applicable to the dispute, the arbitral tribunal can enter the core part of the proceeding, namely the arbitration hearings. During this step of the proceedings arbitrators exercise adjudicative powers and comply with procedural duties. Both elements are related to their activity of deciding principally or incidentally on specific legal issues arising in the dispute. For example the arbitrators have to consider: the admission of evidence, discovery, experts or witness examinations and cross-examinations, requests of preliminary and temporary injunctions, so on and so forth. Depending upon the procedural rules specifically applicable to the dispute, arbitrators’ powers and duties can be subject to wider or stricter limitations. As mentioned above, arbitrators do not have the power called ius imperium. Therefore, assuming the arbitral tribunal deems appropriate to issue a preliminary

27

1. INTRODUCTION

injunction to prevent further damages, it may be necessary to resort to the cooperation of a national court to enforce the arbitral decision. The same holds true also in other circumstances: the arbitral tribunal may not be autonomously able to force a reluctant witness to testify, or to force a non-cooperative party to produce evidence. Most likely, in the latter case arbitrators could play a different card, by threatening the non-cooperative party that they will consider his conduct when making their final decision. However, in general, even if arbitration is private and confidential, the arbitral tribunal may choose, or may be obliged, to request the help of national courts. The basis for the cooperation between the arbitral tribunal and a national court can be provided either by substantive or by procedural rules governing the arbitration procedure which can be very different from case to case. Moreover arbitrators are subject to general duties governing their conduct during the proceedings. The most widely accepted duties include the requirement to treat equally all the parties involved to be impartial and independent in their decisions; to render a valid award; and to apply their best effort to render an enforceable award. This brings our attention to the last step of the arbitration, the granting of an award.

1.2.1.5

The Arbitral Award

The ultimate objective of the arbitral tribunal is to provide an award, namely a final decision on the merits of the case, which corresponds to the judgement of a national court. An arbitral award typically consists of two parts: the reasoning and the operative part. The former part is necessary to guarantee the validity of the award (i.e. the soundness of the decision) explaining and justifying the tribunal’s legal reasoning; the latter part provides the solution of the dispute allocating liabilities and damages. In certain jurisdictions an award includes damages but it could also include a wide range of other remedies, such as permanent injunctive relief, specific performance, amendment of contractual terms or other legal deeds. In practise the arbitral tribunal could decide to split the award in two parts, having a partial and a final award, where the partial award

28

usually allocates the liability and leaves to the final award the quantification of damages. As a result, the partial award can be pronounced shortly after the conclusion of the hearings, allowing the conflicting business parties to benefit from a fast solution to their dispute. Knowing the outcome they could more easily find a settlement about damages and go back to business, even not necessarily waiting for the arbitral tribunal to go through the complicated calculation of damages. Nevertheless, it may be the case that the conflicting parties are not so conciliatory. The arbitral award could be questioned in front of a national court. In fact, as rightly pointed out by Radicati di Brozolo, an international arbitration can be ‘above, or outside of the national legal system until its final product (the arbitral award) lands before the national court’.55 As mentioned earlier, only national courts have the ius imperium, namely the power to directly enforce their decisions using the state’s power of legitimate use of force. In order to be enforced, an arbitral award needs to fall under the control of national courts which are responsible to determine its validity, recognition and/or enforcement. Therefore, without enforcement an arbitral award does not have much more value than a blank page of paper. On the very assumption of the existence of this possibility to second-look review the award, national legal systems consented to the expansion of arbitration and arbitrable fields of law.

1.2.2

The Second-look Review of an Arbitral Award

A wide validity, recognition and enforcement of arbitral awards are vital for the fortune of arbitration as a dispute resolution mechanism. In principle, there is no need to explain why this is a fundamental aspect for the functioning of arbitration. In fact, for example, unless otherwise explicitly agreed by the parties, arbitration is supposed to be a one-stop procedure. There is no appeal and no second instance court. In light of a widely accepted principle arbitral awards should not be subject to a review on the merit of the dispute. Otherwise, if a national judge is allowed to provide his own decision on the merits, the parties’ preferences, believes and expectations (embodied in the arbitration clause) would 55

Radicati di Brozolo [2005].

29

1. INTRODUCTION

be completely neglected. The effectiveness of arbitration as a dispute resolution tool would be nullified and it will be degraded to a costly, time consuming, mock decision with questionable value for any rational business party. The purpose of the New York Convention of 10 June 1958 on the Recognition and Enforcement of Foreign Arbitral Awards (NYC) is to avoid such risk and to guarantee broadest circulation and effectiveness of international arbitral awards as well as to limit the room for national courts’ discretion in reviewing awards. For this reason, the NYC is one of the most successful international conventions of all times, with respect to number of signing countries. In fact, it has been adopted by all EU Member States among with other 142 United Nations Member States. The NYC preserves the effectiveness of international arbitration as a dispute resolution tool but it does not forbid in toto the national courts’ second-look review of an award. An arbitral award can fall under the scrutiny of a national court in three different instances. The losing party may resort to a national court to request the annulment of the award alleging that in light of a breach of the applicable procedural rules the award was not validly pronounced. The winning party may seek: first, the recognition and then, the enforcement of the award in a legal system. The rest of this Section will focus on those three second-look review procedures. 1.2.2.1

Annulment of the Award

Regarding the annulment of the award, both the competence and the jurisdiction on the case belongs to a national court of the state where the parties have chosen to locate the seat of the arbitration. The court will have to verify which are the arbitration rules applicable to the matter; to assess whether those arbitral procedural rules where breached; to decide whether the breach justifies the annulment of the award. Moreover the national court of the arbitration seat is obliged to guarantee that the mandatory rules of its own legal systems where not breached.56 For this reason we have previously argued that the parties choice of arbitration seat is really important. For example, assume that in one country there is a mandatory rule stating that it is against the right of fair trial to ex56

We will discuss later on, in the research, about what are those mandatory rules, the application of which must be guaranteed by the national courts.

30

amine a witness on video-conference. It could be a legal ground for annulment of the award if the parties locate the seat of the arbitration in that country. It is true that, as mentioned above, the arbitrators have the duty to provide a valid award and should comply with those mandatory rules. However, sometimes, arbitrators disregard such rules in good faith, given that it could not always be easily assessed whether a procedural rule of the seat of arbitration is mandatory or whether it could justify the annulment of the award. This is especially true when neither the dispute, nor the arbitral tribunal has any connection with the legal system of the arbitral seat that was chosen only due to its neutrality.

1.2.2.2

The Recognition and Enforcement of the Award

The second procedure of second-look review by national courts occurs during the recognition and enforcement of an arbitral award. While arbitrators may be required to know and comply with applicable arbitration rules as well as with the mandatory rules of the seat of arbitration, they cannot be asked to guarantee the compliance of the award with the rules of each and every possible legal system where the winning party may decide to seek recognition and enforcement of the award. That is why the arbitrators’ duty is to make the best effort to provide an enforceable award. In this respect the New York Convention on the Recognition and Enforcement of Foreign Arbitral Awards (NYC) comes at help. The NYC’s goal is to guarantee a widely recognisable and enforceable arbitral awards strictly defining the exceptions to this default rule. In fact, Article V of the NYC states the limitations to the general principle of recognition and enforcement of an arbitral award. For example indent 2 letter (b) provides that a national court could refuse recognition or enforcement of the award if that will be contrary to the public policy of the state. This research will extensively deal with all the limitations of recognition and enforcement of an arbitral award at a later stage. For the purpose of this Introduction it is sufficient to understand what is the procedure of recognition and enforcement of an arbitral award and its importance for arbitration. This overview runs through the main elements and the main principles gov-

31

1. INTRODUCTION

erning international arbitrations. It clarifies that the starting point and the basis of arbitration lay on the principle of parties’ autonomy which is embodied in the arbitration clause. The step-by-step description of how a typical arbitration proceeds allowed us to identify and discuss the main limitations to the principle of parties’ autonomy and how they have been addressed by different jurisdictions. Also arbitrators powers and duties have been discussed in light of the main goal of any arbitration, namely to solve the dispute through a valid and enforceable award. Eventually, the boundaries to the national courts’ second-look review of an arbitral award have been introduced.

For the sake of intellectual honesty we must admit that this Section and the previous one are subject to substantive simplifications; many existing issues and complications that are at the cornerstone of the arbitration law debate were left out of this overview. This choice reflects the purpose of this Introduction, which is to provide the reader with the working tools necessary for the fruitful consumption of the following Chapters of this research where more advanced issues are discussed.

1.3

Antitrust Arbitration

In this Chapter, we introduced the main characteristics of EU competition law and of international arbitration. As it has been clarified in the problem definition Section of the Preface, this research project focuses on the issues arising from the use of international arbitration as an additional means of the EU competition law private enforcement. In this field of law there are few key issues that remain unresolved, due to a striking legislative gap. Alternative legal rules exist among different legal systems to address these issues. Those rules need to be studied using the Law and Economics methodology indicated in the Preface. This Section introduces the reader to the research sub-questions of this work and it is organised reflecting the structure of this work.

32

1.3.1

Jurisdiction

The first order of questions discussed will be about the jurisdictional issues over the application of EU competition law by an arbitral tribunal. Jurisdiction involves the judicial power of international private adjudicators to resolve disputes that involve a set of mandatory norms, subject to extraterritorial application and, according to the European Court of Justice, constituting European Union public policy rules. Arbitrability of antitrust rules itself is presently undisputed both in the United States and in Europe. The question whether competition law disputes can be adjudicated by way of arbitration received an affirmative answer by the local courts.57 Hence, our attention will firstly focus on how different jurisdictions can determine the validity of a contractual arbitration clause. The two elements that mostly affect the outcome of such evaluation are the following: whose is the competence to define the jurisdiction over the case (whether to the arbitral tribunal or to national judges); according to which legal system the extent of the arbitration clause’s scope should be assessed (i.e. the choice of law issue on arbitrability). Secondly, we will consider issues arising from the fact that competition law rules are subject to extraterritorial application. Such characteristic implies that there may be potentially several different antitrust regimes applicable to the same international dispute. In fact, in a global economy, anticompetitive conducts can easily produce effects outside of the European borders or could be perpetrated by foreign companies abroad, while still affecting the EU internal market. Given that different jurisdictions overlap in international transactions, arbitrators are facing a conflict of law question also on the matter. The consequences to a conflict of law issue may look minor, considering the convergence between different competition law systems. However, in light of the persisting EU competition law exceptionalism in its approach to vertical restraints, the competition regime applied can still determine whether the same conduct will result in two different legal outcomes. Therefore, the applied conflict of law rule is of fundamental importance 57

Blanke and Nazzini [2008a].

33

1. INTRODUCTION

and may justify the parties’ strategic choice of law and/or forum shopping. The legal literature has already shown that practitioners can be influenced in their choice of applicable law by reasons of enforceability instead of merits.58 For instance, if arbitrators expect that the winner will seek enforcement of the award in the US, they may be inclined to systematically choose US antitrust law over European competition law, because it is more likely to produce an award which is enforceable in the US and will not be set aside by the national courts. Thus, for the sake of legal certainty, private parties and arbitrators need an objective and efficient choice of law rule to determine ex ante the competition law regime that is applicable to the dispute regardless of exogenous elements, such as enforcement issues. This objective choice of law rule should also prevent private parties from including into the arbitration clause a choice of law provision that may be driven by the intent to escape from the application of the EU competition law. The above arguments become even more relevant if we consider that the party with more bargaining power may be tempted to favour an arbitration clause which de facto impairs the application of competition law. To summarise, the part of this work dealing with jurisdiction revolves around the extent of validity of an arbitration clause and around the effective solution to the conflict of law issue raised by norms of extraterritorial application. Once these jurisdictional questions are answered, the interaction between competition law and arbitration raises a second order of questions. These issues are particularly challenging and deal with arbitral tribunals’ conduct during the competition law analysis of the case.

1.3.2

Merits

When solving disputes based on EU competition law, the arbitral tribunal must be entitled to a set of powers and duties that are necessary to comply with the enforcement standard required by the public policy rules. The antitrust arbitration issues of merits revolve around the analysis of a set of arbitrators’ powers and duties that must be tailored on the powers and duties referable to 58

Landi and Rogers [2007].

34

national judges. Regarding the arbitrators’ powers, we start from the main question: shall arbitrators have the same powers granted to national judges to solve antitrust disputes? The answer to this question depends on two other fundamental issues: firstly (i), shall arbitrators have full cooperation with national judges and public enforcement? In this case, if Article 16 of Regulation 1/2003 is applicable also to arbitrators, they cannot grant awards conflicting with a decision of the European Commission, but they could also request the assistance of antitrust authorities when deciding a case. However, cooperation between arbitrators and either national courts or the European Commission may raise serious confidentiality issues; it may be incompatible with the arbitration clause; or, in light of the lack of legal provisions on this issue, antitrust authorities may refuse to cooperate with arbitral tribunals. Secondly (ii), should arbitrators have the power to refer for preliminary ruling to the CJEU? According to the current CJEU’s case-law, arbitral tribunals are not entitled to request preliminary rulings to the CJEU under Article 267 TFEU. This may limit the effectiveness of their award, and put at risk the very essence of the arbitrability principle expressed above, which is widely accepted by the legal systems around the world. Whereas regarding arbitrators’ duties, we investigate whether arbitral tribunals shall be subject to the same duties to which national courts are subject as per the application of EU competition law. As mentioned above, arbitrators are already subject to specific duties: to treat equally all the parties involved; to be impartial and independent in their decisions; to render a valid award; and to apply their best effort to render an enforceable award. However arbitrators are private judges and they do not have the function of national judges. On the state of the arts, it is not clear whether they are also obliged to raise competition issues ex officio even when parties forget to do so, or do not want to include such issues in their claim. A negative answer to this question can produce the paradoxical result of an award that has an anticompetitive effect. Moreover, it potentially creates concerns about arbitrators’ personal liability for anticompetitive conduct. Nevertheless, as we will see, a positive answer implies other adverse effects.

35

1. INTRODUCTION

To conclude the Section and this introductory Chapter it is worth reminding that throughout this research project we will apply the Law and Economics methodology to fill the existing legal gap on the application of EU competition law in arbitration proceedings. This gap has allowed legal practitioners to develop contrasting and inefficient solutions to the above mentioned issues. This outcome increases both legal uncertainty and legal costs for business parties. Moreover it jeopardises the functioning of the private enforcement of EU competition law. Our framework of procedural solutions should facilitate the private enforcement and enhance the level of compliance with the EU competition law by market players.

36

Chapter 2 Arbitral Tribunals’ Jurisdictional Issues in the Application of EU Competition Law As mentioned in the introductory Chapter, from a procedural and logical point of view, an adjudicator has to solve all jurisdictional issues, before addressing the merits of the case. This is done at the very beginning of either arbitration or litigation proceedings. The present work follows the same logic. This Chapter and the next one deal with the matters of jurisdiction, while the last two Chapters delve into arbitrators’ powers and duties to assess the merits of the case. In antitrust arbitration, legal practitioners have focused their attention upon three key issues of jurisdiction. The answers given to these questions diverge among legal systems and produce different outcomes. Hence, economic theories can provide some guidance, comparing the alternative solutions and displaying the existence of latent inefficiencies. The first and the second issues pertain to whether the arbitral tribunal is entitled to solve the case at stake. These two issues are the two sides of the same coin which is called arbitrability. They are based on the two conditions that need to be satisfied in order for the arbitral tribunal to have jurisdiction over the case:

37

2. ARBITRAL TRIBUNALS’ JURISDICTIONAL ISSUES IN THE APPLICATION OF EU COMPETITION LAW

1. it should be possible, in principle, to subject to arbitration the specific dispute at stake, the so-called objective arbitrability; and 2. the specific dispute should fall within the scope of a valid contractual arbitration clause upon which the parties have ex ante negotiated and agreed, we can call the latter subjective arbitrability.

The third issue which draws the attention of legal practitioners deals with the choice of law rule applicable to the dispute. In fact, once it is ascertained that the dispute can be solved in arbitration, the arbitral tribunal has to identify which law is applicable to the case. All these three issues, that will be discussed in details below, belong to the matters of jurisdiction understood in a broad sense.1 Furthermore, they deserve to be analysed in the same Chapter because they all face the same fundamental problem from different perspectives. Different legal systems provide various rules to identify the adjudicative body entitled to answer the issues above. It can be either arbitral tribunals or national courts. In each legal system, the chosen adjudicative body may have the duty or the incentive to apply different and incompatible rules or criteria to address the three jurisdictional issues mentioned above. While this scenario is not per se problematic in international arbitration, it can have dangerous effects in the field of antitrust arbitration. The answer given to each one of the three questions described above can produce diametrically opposite outcomes in antitrust disputes, which are detrimental to the interests of the parties and to international commerce, due to the resulting legal uncertainty and unpredictability. In other words, the core topic of this Chapter revolves around rules used to define the adjudicators’ competence as well as the choice of 1

The third question is considered a procedural matter in some legal systems and an issue of merits in others. Therefore this latter topic will be discussed and used as a bridge to carry the reader from this Chapter on jurisdiction to the following Chapters on the merits.

38

law problem in international transactions. On the one hand, defining the adjudicators’ competence means to identify the rules that determine which is the adjudicative body entitled to address the three questions described above. Namely, whether it will be an arbitral tribunal or a national court to decide upon the arbitrability and choice of law issues. Under the principle of kompetenz-kompetenz, described in the introductory Chapter, arbitrators can decide upon their competence. However antitrust law is composed of mandatory rules of public policy nature that can contrast with the principle of kompetenz-kompetenz. On the other hand, the unresolved topic known as choice of law has been part of the legal scholars’ attention for centuries now.2 This topic is even more relevant in the present global economy. International transactions imply the application of contrasting mandatory rules that may overlap due to their extraterritorial application, as in the case of antitrust law. In fact, in such cases, legal systems hardly tolerate the use of either forum shopping or choice of law clauses. Such practises may limit the application of their mandatory rules and impair their ability to achieve the goals pursued through their policies. Both these topics revolve around the effectiveness of arbitration as an Alternative Dispute Resolution mechanism and as a tool for the private enforcement of competition law, hence they are fundamental elements for the predictability and the development of international commerce. To summarise, this Chapter addresses the three main jurisdictional issues faced in international arbitration and mentioned above, showing how the alternative rules adopted by legal systems affect individuals’ incentives to arbitrate, increasing the costs of the dispute resolutions. The goal of this Chapter is to provide a satisfactory solution which will increase the effectiveness of arbitration as an antitrust private enforcement mechanism. This Chapter gives a thorough description of the legal debate about each of the three issues introduced above; it 2

Ruhl [2006].

39

2. ARBITRAL TRIBUNALS’ JURISDICTIONAL ISSUES IN THE APPLICATION OF EU COMPETITION LAW

outlines the characteristics of these problems, relevant for the economic analysis; then it provides an overview of the alternative rules used in similar cases. The next Chapter builds upon this description and develops our economic analysis based on game theory, to show the inefficiencies of the current framework and to provide possible efficiency-increasing solutions.

2.1

The Legal Debate on Antitrust Arbitration Clauses

2.1.1

The Objective Arbitrability of Competition Law

As mentioned in the introductory Chapter, private parties can agree to solve their disputes in front of arbitral tribunals instead of resorting to national courts on the basis of the principle of party autonomy. Nevertheless, the outcome of an arbitration procedure may affect rights and interests that go beyond those of the two disputants. While national judges are public bodies, international arbitrators are private individuals serving the interest of the conflicting parties; they do not retain obligations towards third-parties or public interests. Hence, every legal system defines limitations to the principle of party autonomy in arbitration through the concept of arbitrability.3 Denying the arbitrability of specific rights, the lawmaker decides that arbitral tribunals are not a suitable “forum” to dispute these rights. Denying the objective arbitrability, the lawmaker obliges private parties to enforce specific rights under the whole framework of guarantees embodied in state procedural rules. In other words, the second-look review, in the phase of recognition or enforcement of an arbitral award, is not considered enough to guarantee the protection of third-parties or state interests related to these specific rights. Therefore, it can be said that when the lawmaker denies the objective arbitrability of certain rights, it is using an extreme measure of protection of those rights, forbidding in toto the possibility for arbitrators to solve any dispute based on those rights. 3

See introductory Chapter, above.

40

At the international level, among different legal systems, there is no uniform definition of such limitations to arbitrability. This legal gap determines inconsistencies upon the definition of the rights, or fields of law, that are not arbitrable in the objective sense and that should be adjudicated only by national courts. For example, in France it is not possible to arbitrate issues related to personal status and divorce;4 in England, it is not possible to arbitrate disputes based on patents granting;5 in China, it is not allowed to arbitrate disputes with the public administration;6 and so on and so forth. Regarding antitrust law, although the objective arbitrability of rights based on competition law has been widely recognised in Western legal systems, it is not universally and explicitly accepted.7 As will be described below, there are pockets of uncertainty as far as emerging economies are concerned. Therefore, it can be understood why the arbitrability of competition law falls within the choice of law problem addressed in this Chapter. It is out of the scope of this research to discuss whether antitrust arbitration threatens public interest to such an extent to justify the denial of competition law arbitrability. However, in order to clarify why arbitrability of competition law can still be an issue, it will be useful to retrace the main arguments used in favour or against antitrust arbitrability. These arguments retrace the historical evolution of Western legal systems, towards a wide acceptance of the abstract arbitrability of competition law.

2.1.1.1

The American Safety Doctrine Denies Antitrust Arbitrability

The objective arbitrability of rights based on antitrust law was first affirmed by the US Supreme Court in the 1985 ground-breaking decision known as the Mitsubishi case.8 Before this 1985 case, the dominant theory on antitrust arbitrability was derived from the American Safety doctrine called after a case decided in 4

See Article 2060 of the French Civil Code. Mustill and Boyd [1989], p. 146. 6 Article 3 of the Law issued on 31 August 1994. 7 Blanke and Nazzini [2008a]. 8 Mitsubishi Motors v. Solar Chrysler-Plymouth, 473 U.S. 614, 105 S. Ct. 3346 (1985). 5

41

2. ARBITRAL TRIBUNALS’ JURISDICTIONAL ISSUES IN THE APPLICATION OF EU COMPETITION LAW

1968.9 In practise, this doctrine stated that private parties do not have the free disposition of antitrust rights. Those rights are embodied in mandatory norms excluded from the general rule of arbitrability under what is called the public policy judicial exemption.10 The aversion to arbitration was based on the fear that the Alternative Dispute Resolution procedures would be used to circumvent the application of the US antitrust law and impair the effectiveness of its policies. Four main arguments were used by the court in the American Safety case to support this decision.11 • Antitrust claims are inappropriate for enforcement through arbitration because of the public interests and public nature embodied in this field of law. ‘[A] claim under the antitrust laws is not merely a private matter. Antitrust violations can affect hundreds of thousands, perhaps millions, of people and inflict staggering economic damage.’12 Moreover, given the fact that arbitrators are often members of the business community, they are not suitable for applying the legal rules meant to limit and control the very same business community. • Arbitration could limit and impair the effectiveness of national courts private enforcement of antitrust law. It is worth noting that historically in the US private enforcement of antitrust law in front of national courts has been the main and fundamental tool for application of antitrust rules. The reason of its success is strongly based on the possibility for the victim to obtain treble damages.13 In the viewpoint of the judges that pronounced the American Safety Doctrine, since treble damages are a sanction that has a 9

American Safety Equipment Corp v. J.P. Maguire and Co., 391 F.2nd 821, 825 (2d Cir. 1968). 10 American Safety Equipment Corp v. J.P. Maguire and Co., 391 F.2nd 821, 825 (2d Cir. 1968). 11 Ironically the same old arguments are, still used by opponents of competition law arbitrability. 12 American Safety Equipment Corp v. J.P. Maguire and Co., 391 F.2nd 821, 825 (2d Cir. 1968). 13 For the sake of completeness, should be noted that the success of private enforcement in the US is also explained by the presence in that legal system of class actions and the possibility of contingency fees. See Zekos [2008b], p. 19.

42

penal nature, arbitrators cannot grant this remedy and parties would have escaped the application of this deterrent tool using arbitration. • It has also been argued that contractual arbitration clauses would be unilaterally dictated by the economically stronger contractual party, who will draft the clause in such a way to impair the effectiveness of antitrust rules. A similar argument has been used also against arbitrability in the fields of consumer protection or employment law. • Lastly, in the American Safety case, the court believed that the arbitration proceeding is not a suitable tool to solve antitrust disputes, because of its flexibility, celerity and simplicity which contrast both with the high complexity of antitrust cases and with the problematic fact-finding, typical of this legal field.14 It is interesting to note that the arguments used in the American Safety doctrine are contrasting and inconsistent with another contemporary widespread position. In fact, both former case-law and legal scholars agreed upon the possibility for private parties to settle the very same antitrust disputes out of court.15 Public interests related to antitrust law are not protected during private settlement agreements occurring out of courts.16 Also all the other American Safety arguments cannot be reconciled with antitrust settlements. 2.1.1.2

The Breakthrough Mitsubishi Decision

In the 1985 Mitsubishi case, the US Supreme Court overcomes each and every one of the arguments used to support the American Safety doctrine and admits the arbitrability of US antitrust law.17 At this point it is useful to overview the main facts defining the Mitsubishi case. 14

American Safety Equipment Corp v. J.P. Maguire and Co., 391 F.2nd 821, 825 (2d Cir. 1968). 15 Johnson [1983] pp. 99-101. 16 In fact, reaching a settlement, the private parties could find a way to share the profits of an anticompetitive agreement which still harms public interests. 17 Mitsubishi Motors v. Solar Chrysler-Plymouth, 473 U.S. 614, 105 S. Ct. 3346 (1985).

43

2. ARBITRAL TRIBUNALS’ JURISDICTIONAL ISSUES IN THE APPLICATION OF EU COMPETITION LAW

Mitsubishi is a Japanese car manufacturer who signs along with Chrysler a vehicles distribution agreement with Solar, a reseller incorporated in Puerto Rico.18 Within the distribution agreement, the parties agreed through a choice of law clause that the law applicable to the contract was the Swiss law. The contract also included an arbitration clause stating that all disputes between the parties should be solved through an arbitration proceeding in Tokyo, Japan, under the procedural rules of the Japan Commercial Arbitration Association. The dispute is based on a clause which requires from Solar, before exporting cars to Central and South America as well as to some parts of the US, to obtain the specific authorisation from Mitsubishi. Such territorial limitations were not the only vertical restraints included in the distribution agreement which could be challenged under competition law. The contract also provided minimum purchase quantities which Solar was not able to resell unless provided with the authorisation to export in the territories mentioned above. When Mitsubishi denied the authorisation, Solar refused to accept further deliveries breaching the contractual clause about minimum purchase quantities. After Solar’s refusal to participate to the arbitration procedure in Japan, the parties started a litigation in front of the first instance court of Puerto Rico. While Mitsubishi tried to enforce the arbitration clause, requesting an order to compel to the arbitration clause from the local court, Solar claimed that Mitsubishi’s conduct was in violation of the US antitrust law and sought compensation and damages. The court of first instance declares that the jurisdiction on this dispute belongs to the arbitral tribunal in Japan. However the court of appeal reverses the outcome applying the American Safety doctrine. The court of appeal states that this doctrine covers also international business transactions, therefore it is forbidden to solve in arbitration disputes based upon the application of US antitrust law. In July 1985, when the US Supreme Court had to decide the case, it stated that antitrust matters are arbitrable finally overruling 18

It is worth reminding that Puerto Rico is a territory falling within the jurisdiction of the US legal system.

44

the American Safety doctrine.19 The US Supreme Court was substantially asked to provide guidance on how to address the trade-off between the needs of public policy as well as of international trade. The answer is not surprising considering that during the 80’s, the court was acting on the background of strong pressure from both political and economic liberalism.20 In the Mitsubishi case, the reasoning of the US Supreme Court builds upon two principles expressed by previous case-law:21 • in the context of international business transactions, the parties’ choice of forum clause is to be considered binding and to be respected by the US courts; • agreements aiming to define ex ante the law applicable and the forum where disputes will be solved are fundamental to guarantee legal certainty and predictability. In the Mitsubishi case, the US Supreme Court takes these two principles even further. Notwithstanding the importance of the application of mandatory antitrust rules, the US Supreme Court considers that parties’ autonomy to choose international arbitration as Alternative Dispute Resolution mechanism is indispensable for the development of international business trade.22 Therefore, in this trade-off the judges favoured the side which is in line with the needs and the principles embodied in the process of market globalisation. In the words of the US Supreme Court: ‘We cannot have trade and commerce in world markets and international waters exclusively on our terms, governed by our laws and resolved in our courts’.23 19

More specifically the Mitsubishi case refers exclusively to cases of international transactions. Arbitrability of domestic disputes on antitrust matters comes later with the cases Shearson/American Express Inc. v. Mcmahon, 482 U.S. 220 (1987) and Rodrigues de Quijas v. Shearson/Lehman Bros., 845 F.2d 1296 (5th Cir. 1988). 20 Frieden and Kennedy [2006], chapter 16. 21 The Bremen v. Zapata Off Shore Co., 407 U.S. 1 (1972) and Scherk v. Albero-Culver, 417 U.S. 506 (1974). 22 Serravalle [1986]. 23 Mitsubishi Motors v. Solar Chrysler-Plymouth, 473 U.S. 614, 105 S. Ct. 3346 (1985), par. 5074.

45

2. ARBITRAL TRIBUNALS’ JURISDICTIONAL ISSUES IN THE APPLICATION OF EU COMPETITION LAW

This approach is diametrically opposite with respect to the reasoning of the American Safety doctrine. In fact, the Supreme Court takes into consideration and systematically demolishes each and every one of the arguments expressed by the American Safety doctrine.24 More specifically, the Court focuses on the relationship between arbitration and the protection of public interests as key function of antitrust law. The Judges state that this field of law, on the basis of its own nature, cannot be excluded from parties autonomy. Within the context of court litigation, private parties already have the disposition of rights based on antitrust law. Most certainly, private enforcement mechanisms, contrarily to public enforcement ones, leave the parties free to decide whether or not to start a legal proceedings against anticompetitive conducts, and to interrupt litigation by settling the case almost at any time.25 Therefore it is inconsistent and incorrect to exclude parties’ autonomy to arbitrate disputes involving rights based on antitrust law. With the Mitsubishi decision the objective arbitrability of disputes based on antitrust law matters is unequivocally ascertained in the US. However, the existence of the trade-off which involves the public policy is not disregarded. The US Supreme Court clarifies how the guarantees therein embodied can be safeguarded during the second-look review procedure of the arbitral award in front of national courts. Such procedure represent an ex post guarantee that the main principles embodied in the US antitrust law are not breached, overlooked or ignored. The side-effect of leaving such wide back-door open for national judges’ interferences in arbitration creates a brand new set of legal issues and uncertainties which are discussed below. In the very same decision there is even an example of public policy breach that should be censured by the national court. It is contained in the famous footnote 19 by the US Supreme Court in the Mitsubishi case.26 ‘In the event of the choice-of-forum and choice-of-law clauses operated in tandem as a prospec24

Mitsubishi Motors v. Solar Chrysler-Plymouth, 473 U.S. 614, 105 S. Ct. 3346 (1985). Bernardini [1985]. 26 Mitsubishi Motors v. Solar Chrysler-Plymouth, 473 U.S. 614, 105 S. Ct. 3346 (1985) par. 637. 25

46

tive waiver of a party’s right to pursue statutory remedies for antitrust violations, we would have little hesitation in condemning the agreement as against public policy’. In practise the wide meaning and the uncertainties resulting from this kind of clauses open a “pandora box” that could potentially jeopardise the effectiveness of international arbitration.27 Notwithstanding the above, the Mitsubishi case is still the turning point in the application of competition law in arbitration. The objective arbitrability of antitrust law is explicitly affirmed and included into a legal system for the first time.

2.1.1.3

The Objective Arbitrability of Competition Law in Europe

The arbitrability of competition law in Europe is ascertained following a different route in comparison to the US experience. Even if in Europe there is still no explicit legal provision on this point, the same result has been indirectly achieved as common practise. Regarding the arbitrability of the EU competition law, the European Commission (hereinafter the Commission) was the first institution to show a positive attitude towards this Alternative Dispute Resolution mechanism. Then it was the turn of the Court of Justice of the European Union, through its case law, to take a position in favour of the arbitrability of the EU competition law. Already in the 70’s, the Commission subordinated several of its exemption decisions, under Article 101(3) TFEU, to the requirement of being informed by the beneficiaries of the exemption about any future arbitration proceedings related 27

This footnote means that national courts should censure arbitral clauses which have the intent or the effect, in theory or in practise, to deprive the parties from their rights to seek remedies against anticompetitive behaviours in front of an arbitral tribunal. This argument pointed out by Mourre (Arbitrability of Antitrust law in the US a EU Perspective in Blanke and Landolt [2010]) rises a set of further questions: is the arbitral clause itself to be unenforceable shifting the jurisdiction back to national courts or is the arbitration clause valid and only the provisions limiting the arbitrators’ powers to be void? Moreover, which is the relationship between the principle of “footnote 19” and the parties’ autonomy? Those questions lead to the second set of issues discussed in the following paragraphs.

47

2. ARBITRAL TRIBUNALS’ JURISDICTIONAL ISSUES IN THE APPLICATION OF EU COMPETITION LAW

to the exemption.28 It can be argued that, even if the Commission could have denied the arbitrability of disputes based on those decisions, by trying to assure that the EU competition law is not breached in arbitration proceedings, in principle, the Commission has implicitly recognised the arbitrability of this field of law. Obviously this argument is not conclusive nor remotely sufficient to state the Commission’s positive attitude towards arbitration. However its positions became more clear first within the Commission’s Regulation 556/89, then within the Commission’s Regulation 1400/2002. Specifically in the latter, which covers block exemptions for vertical agreements in the automotive market, in Article 3 number 6, the Commission prescribes that the exemption of the vertical agreement requires the inclusion of a clause granting to both parties the right to appeal to an expert or an independent arbitrator to solve future disputes on the performance of the contract.29 Also in the Commission’s UIP decision dated 12 July 1989, which authorises a common distribution agreement between three American movie companies affecting the European market, the Commission imposed to the involved parties the inclusion of an arbitration clause to solve any future dispute on the contract. Although also these examples do not contain an explicit nor general confirmation of the arbitrability of competition law matters, it can be stated that in the eyes of the Commission the arbitrability of the EU competition law has been in principle accepted.30 Focusing now on the case-law of the CJEU, there are three key decisions that first clarified the CJEU’s position on the arbitrability of the EU competition law. The three cases Nordsea,31 Municipality of Almelo 32 and Eco Swiss 33 all pertain to the application of the EU competition law in arbitration proceedings. In fact, 28

Henkel/Colgate 26/1/1972; Bayer/Gist 15/12/1975; 5/2/1976; Kabelmetal 18/7/1975; 22/8/1975; De Laval 25/7/1977, 22/8/1977; Rockwell/Iveco 13/7/1983, 17/8/1983. 29 However it was also required that each party could appeal to a state court. Therefore such clauses cannot be considered as binding arbitration clauses. 30 Lugard [1998], p. 295 31 Case C-102/81, Nordsee Deutsche Hochseefischerei GmbH v Reederei Mond Hochseefischerei Nordstern AG & Co KG [1982] ECR 1095. 32 Case C-393/92 Gemeente Almelo et al. v. NV Energiebedrijf IJsselmij [1994] ECR I-1277. 33 Case C-126/97, Eco Swiss China Time Ltd v Benetton International NV [1999] I-3055.

48

in the first two cases the CJEU deals with the relationship between arbitral tribunals and Article 267 TFEU, namely whether arbitral tribunals which apply the EU competition law are entitled to request a preliminary ruling from the CJEU. This topic will be discussed in Chapter 4, for now it is enough to recall that according to the CJEU arbitral tribunals do not fulfil the requirements of Article 267 TFEU, hence they do not have the power to use the referral procedure therein provided. In the third decision, namely the Eco Swiss case, the CJEU had to define national court’s powers to review the application of EU competition law by arbitrators during the second-look review of the award.34 Also this topic will be discussed at a later stage of this work, for now it suffices to remind that the CJEU leaves a wide and undetermined power of review in the hands of national courts. Although, as it is clear from the above, the three cases pertain to the application of EU competition law in arbitration, none of them addresses the arbitrability as a central issue of the dispute nor it is a specific question discussed by the CJEU. Nevertheless, there is no doubt that the arbitrability issue is an implied question which logically needs to be preliminary solved by the CJEU before even starting to discuss the specific topics of each preliminary ruling request. It is undisputed that there is no other way to explain or justify the CJEU’s silence on the arbitrability, unless admitting that the CJEU did not touch upon this point since it is well established and clearly accepted.35 The Eco Swiss case is the most interesting in this respect, since the CJEU explicitly discusses the public policy role of the EU competition law rules as well as the need to guarantee that the application of these mandatory norms is not avoided by the parties through arbitration. The CJEU clearly states that: ‘Article 85 [now Article 101 TFEU] of the Treaty constitutes a fundamental provision which is essential for the accomplishment of the tasks entrusted to the Community and, in particular, for the functioning of the internal market’.36 In the eyes of the CJEU, the above argument along with 34

Case C-126/97, Eco Swiss China Time Ltd v Benetton International NV [1999] I-3055. Blanke and Nazzini [2008a]; Radicati di Brozolo [1999], pp. 665-697; Bastianon [1999], pp. 471-483. 36 Case C-126/97, Eco Swiss China Time Ltd v Benetton International NV [1999] I-3055, Par. 36. 35

49

2. ARBITRAL TRIBUNALS’ JURISDICTIONAL ISSUES IN THE APPLICATION OF EU COMPETITION LAW

the fact that under Article 101(2) any agreement or decision prohibited pursuant to that Article is automatically void are sufficient to claim that ‘the provision of Article 85 [now Article 101 TFEU] of the Treaty may be regarded as a matter of public policy within the meaning of the New York Convention.’37 Furthermore the CJEU does not oppose the application of competition law in arbitration but only provides, later in the decisions, that it is for the national courts, during the review of the award, to guarantee that the EU law provisions are applied by arbitrators according to an uniform interpretation.38 As noted by Radicati di Brozolo, the fact that in all these occasions the CJEU have never even touched upon the issue of arbitrability strongly reveals that the CJEU agrees with the majority of the legal literature admitting the arbitrability of competition law cases.39 To summarise, both the case-law and the practise of EU institutions confirm definitively that under the EU law it is not forbidden to arbitrate disputes involving the application of EU competition law. This means that ultimately the choice on this procedural issue is in the hands of the EU Member States. The EU Member States have widely accepted the arbitrability of the EU Competition law. Although it was not necessary, the Swedish Arbitration Law issued in 1999 explicitly mentions competition law in its Section I stating that: ‘Arbitrators may rule on the civil law effects of competition law as between the parties’. Other legal systems simply accept it as granted, without any need to mention or discuss the point. When it these jurisdiction the issue was raised in front of a national court, the latter has simply ruled accordingly. In France the arbitrability of competition law was affirmed in numerous cases.40 In Italy as suggested by the Court of Appeal of Milan, ‘any doubts [about the arbitrability of competition law] are now overcome both by the legal literature and by the 37

Court of Justice of the European Union, Case C-126/97 on 1 June 1997, par. 39. Court of Justice of the European Union, Case C-126/97 on 1 June 1997, par. 40. 39 Radicati di Brozolo [1999], pp. 665-697. 40 Court of Appeal of Paris in Almira Films vs. Pierrer (1989); CA Paris, 29 March 1991, Ganz v. Socit Nationale des Chemin de Fers Tunisiens, Rev. Arb. [1991], 478, note L. Idot; CA Paris, 19 May 1993, Labinal v Mors, Rev. Arb. [1993], 645, note Ch. Jarrosson,JDI (Clunnet)[1993], 957, note L. Idot.; CA Paris, 14 October 1993, Aplix v. Velcro, Rev. Arb. [1994], 164, note Ch. Jarrosson. 38

50

case-law, either at the national and the communitarian level’.41 Furthermore the Court of Appeal of Milan specifies that: ‘The concepts of “freely disposable” rights and that of mandatory rules should not be confused or mixed, as not all mandatory rules or rules of public policy relate to rights that are not in the free disposition of the parties... Antitrust rules are mandatory rules, which are part of the concept of “economic public policy”, but they do not relate to rights that the parties cannot freely dispose of... As a consequence, as a matter of principles, issues of competition law, be they of Italian or of Community law, can be deferred to arbitrators, and there is no difference from that perspective between domestic and international arbitration.’ The issue of arbitrability is not even considered as a questionable matter in Germany,42 England,43 and the Netherlands.44 The topic has fallen also under the consideration of several arbitral tribunals, which have discussed it into their final awards, solving positively the issue on the arbitrability of competition law.45

2.1.1.4

The Conflict of Law on the Antitrust Objective Arbitrability

In light of the above, we can state that in the US and the EU legal systems the arbitrability of competition law is generally accepted and undisputed. However, the fact that in general the objective arbitrability of competition law is pacifically recognised does not mean that the arbitrability issues lose their practical relevance for the private enforcement of competition law in arbitration. Even if in Europe competition law is arbitrable, there are still differences in defining objective arbitrability that can affect the private enforcement in arbitration. A simple example can help to clarify the problem underneath the scenario described above. Under the 1961 Belgian regulation concerning the unilateral 41

Court of Appeal of Milan, Terrarmata Tensacciai, 2006. Landolt [2006], page 95. 43 Blanke and Nazzini [2008a], p. 50. 44 Poudret and Besson [2007], p. 297. 45 International Chamber of Commerce (ICC) Award no. 1397 of 1966; ICC Award no. 2811 of 1979; ICC Awar no. 7097 of 1993; ICC Award no. 7673 of 1993, ICC Award no. 7539 of 1995. 42

51

2. ARBITRAL TRIBUNALS’ JURISDICTIONAL ISSUES IN THE APPLICATION OF EU COMPETITION LAW

termination of exclusive distribution agreements, arbitral tribunals do not have jurisdiction to solve disputes based on this mandatory law.46 However, distribution agreements do typically include an arbitration clause and they may potentially produce disputes based on the application of competition law. This means that disputes over those specific contracts involving competition law are not arbitrable in Belgium, while they are arbitrable elsewhere in Europe and in many other jurisdictions in the world. When either international arbitrators or national courts (intervening ex ante or during the second-look review) have to determine whether the arbitral tribunal has jurisdiction on such a dispute, the law applicable to determine the objective arbitrability becomes the key issue. If the arbitrability has to be decided applying the Belgian 1961 Act, the arbitrators or the national court will have to deny the arbitral tribunal’s jurisdiction over the case. Otherwise, under the case law of Belgian courts, this is a public policy issue which justifies the annulment of an award and its deny of recognition or enforcement under the NYC.47 Alternatively, the adjudicator can decide that, while the Belgian 1961 Act is indeed the law governing the merit of the case, another law is applicable to the issue of arbitrability and the arbitral tribunal has jurisdictions on the case. Therefore, when the objective arbitrability of a dispute is assessed applying the Belgian law, the effectiveness of arbitration as antitrust enforcement mechanism can be questioned. This simple example identifies the problem discussed in this Chapter. Different legal systems provide various limitations to the concept of objective arbitrability, trying to guarantee that the public interests involved in a private dispute are protected. In disputes with transnational elements the outcome of the objective arbitrability issue is determined and affected by both the identity of the authority entitled to solve the conflict of law matter and by the rule adopted to decide the conflict of law matter. In different legal systems, either the national court or the arbitral tribunal, can be in charge of deciding upon the issue of arbitrability. As explained below, these two adjudicators will use different standards 46 47

Zekos [2008a], p. 367. Zekos [2008a], p. 367.

52

to determine the law applicable to address this issue. Therefore the key issue becomes the criteria used to decide which is the law applicable to the question of arbitrability. In the example above we used a well known arbitrability limitation of a European Member State’s legal system, but it is easy to understand the degree of uncertainty that emerges when the legal system of developing countries are involved. In the presently global markets, this debate cannot be limited to the EU and the US, the two legal systems with the most advanced legislation and case-law on competition law. The discussion has to take also into account the antitrust enforcement approach followed by other legal systems. Regarding objective arbitrability, we are not aware of any legal system which explicitly denies the arbitrability of competition law. However in many jurisdictions there is a legal gap on the issue; and in the best case scenario there are only few cases decided by lower courts. Hence, this is all but a solved matter. Mostly, this is an open question in those jurisdictions which are not highly sensible either to international arbitration or to competition law enforcement. As an example of the different approach followed by developing economies, we can focus on the Russian legal system. Russia has joined the WTO in 2012 and is the third economic partner of the European Union, after the US and China, but most importantly it is a country which for geopolitical and historical reasons trades with European counterparties through big monopolists in the sector of energy production (i.e. Gazprom). It can be argued that in Soviet Russia international commercial arbitration was an uncommon tool for dispute resolution, mostly due to the fact that the Government was controlling every aspect of international trade of the country. The Russian Federation has a relatively recent experience with international arbitration. The national leading arbitration institution, the International Commercial Arbitration Court at the Chamber of Commerce and Industry of the Russian Federation (ICAC) (International Commercial Arbitration Court at the Chamber of Commerce and Industry of the Russian Federation), celebrated its 20th anniversary on 11th June 2013. In 2012 the ICAC registered 241 claims, 32% of which included a European party, while the 64% pertained to the field

53

2. ARBITRAL TRIBUNALS’ JURISDICTIONAL ISSUES IN THE APPLICATION OF EU COMPETITION LAW

of international sale of goods.48 However, as stated by the famous Russian legal practitioner Alexander Khrenov, ‘the Russian Federation is not perceived to fall within the pro-arbitration jurisdictions category as far as recognition and enforcement of foreign arbitral awards are concerned’. As a matter of fact, an independent study resulting from a questionnaire filled by 136 respondents, described as corporate counsels at leading corporations around the world, 67 of which have been also in-depth interviewed, tells us that this forum is not so popular among business parties.49 The report suggests that the choice of arbitration seat is largely affected by the ‘formal legal infrastructure’ of that country, then by the law governing the contract and also by general convenience of the location including additional elements, such as proximity to the business. While London gains the role of most frequently chosen arbitration seat, the respondents of the survey have the most negative perception of Moscow and mainland China. Nevertheless, the 2013 survey, released by the School of International Arbitration, by the Centre for Commercial Law Studies of the Queen Mary University of London, provides empirical evidence that in some key sectors where Russian companies play a relevant role, the use and the preference of arbitration are solidly confirmed by business parties. The survey shows that arbitration is more popular in some industry sectors than others. Most notably, among the surveyed companies, the 78% in energy, the 84% in construction and the 69% in the financial services strongly claim that international arbitration is well suited to resolve disputes in their industry.50 In light of this data, what is relevant for the present analysis is whether Russia related disputes on energy, construction and financial services solved in arbitration can be based on competition law issues. Namely, whether antitrust arbitration is objectively arbitrable under the Russian law. There is no direct answer to this question within the Russian federal legislation. Although Russia has signed and ratified the NYC, in the NYC there is only a general reference to the concept of arbitrability and of public policy matters which are not defined. The 48

The official data is available on the website of the ICAC (International Commercial Arbitration Court at the Chamber of Commerce and Industry of the Russian Federation) 49 launches in Vancouver. 50 PWC and Queen Mary University of London [2013].

54

definition of what kind of disputes can be solved by arbitrators remains within the sphere of power of the signatory States of the NYC. The Russian sources of law, which address this matter, do not provide any guidance about the arbitrability of antitrust disputes. In fact, the Russian Federal Law on the International Commercial Arbitration of 1st July 1993 No. 5338-1 (ICA Law) provides at its Article 1(2) the following definition of arbitrability: ‘In international commercial arbitrations, upon the resulting agreement of the parties, it is possible to devolve to arbitrators disputes originated from contract law matters or civil law issues.’51 At indent number 4 of the same article there is a specification: ‘This law does not affect what is provided in any other existing Russian law according to which specific disputes can not be arbitrated or that can be arbitrated only according to other directives which are not included within this law.’52 This means that the non-arbitrability could be included, for instance, in a provision of the Russian federal competition law. However, in practise, neither the latter nor any other Russian law limit the antitrust jurisdiction of international arbitrators or include an explicit prohibition to the antitrust arbitrability. The lack of any explicit limitation is contrasting with the Russian legal literature which claims that those matters are not arbitrable.53 The legal literature justifies its position against antitrust arbitrability arguing that international arbitral tribunals could solve the dispute avoiding the application of Russian competition; and this is in contrast with the mandatory law nature of competition rules. Until the present day, the Russian Supreme Arbitrazh (Commercial) Court has not faced any case on this issue. However there is case-law of other lower Russian courts which had to address such question of arbitrability. Although the number of cases is limited, that allows to hypothesise the possible approach and the attitude of Russian courts on this matter. One of the examples is the resolution of the North-Western Circuit of the FCC.54 The dispute as between Borregaard Indastries Ltd. v. OAO Vyborskaya 51

Translated from Russian by the Author. Translated from Russian by the author. 53 Gurkov [2013]. 54 Federal Arbitrazh (Commercial) Court of the north-west circumscription case decided on 10 may 2011, case n. A56-68936/2010. 52

55

2. ARBITRAL TRIBUNALS’ JURISDICTIONAL ISSUES IN THE APPLICATION OF EU COMPETITION LAW

Tsellyuloza, the two parties signed a distribution agreement including an arbitration clause. Borregaard started the arbitration proceeding and the arbitral tribunal composed by the Arbitration institute of OSLO solved the dispute in favour of the claimant (Borregaard), who demanded enforcement of the award in front of the Arbitrazh (Commercial) Court of the Saint Petersburg and Leningrad Region. The Court enforced the arbitral award according to the claimant request. The defendant appealed the decision to the Federal Arbitrazh (Commercial) Court of the North-Western Circuit. In its decision the federal appeal court noticed that the arbitrators have been obliged to apply provisions of competition law to solve the case. At the same time the court did not rise the issue about whether such kind of disputes can or cannot be subject to arbitration. Neither the Appeal court did verify the correct application of competition law by the arbitral tribunal. It decided to confirm the first instance court’s decision without any amendments. Therefore in this case, the appeal court indirectly acknowledged that the disputes based on the application of competition law can be arbitrated. The arbitrability of antitrust disputes is also recognised in the same way in some “decrees” of other lower Arbitrazh (Commercial) Courts. Nevertheless it can be argued that uncertainty is persistent. In fact, neither higher courts nor the law clarify explicitly whether antitrust law is arbitrable. On the one hand, the legal literature solves this gap denying arbitrability while, on the other hand, there are few instances of case-law which interpret the same sources of law as allowing the arbitrability of antitrust case. This case-law is not decisive, since in the Russian legal system neither the former nor the latter interpretation is binding upon courts which will be addressing the same issue in future cases. Until the Russian Supreme Arbitrazh (Commercial) Court does not solve a case dealing with such issues the uncertainty will persist. Therefore, the Russian courts have the right to interpret and apply the above mentioned provisions of law either pro or against the arbitrability of competition law matters. This analysis of the Russian scenario and the previously described Belgian scenario highlight the two sources of uncertainty affecting the issue of objective arbitrability of competition law:

56

• Firstly, who is the adjudicator entitled to decide on the issue of arbitrability. More specifically who has to solve the potential conflict of law between competing rules on the question of arbitrability; • Secondly, what is the conflict of law rule that should be used to decide the applicable law to the question of arbitrability. The first source of uncertainty comes from the possibility of forum shopping between arbitrators and national courts, as well as among various national courts in different jurisdictions. The next Section will be focused on this point. The analysis here starts from the second source of uncertainty, which comes from the different rules used to solve the issue of objective arbitrability. An overview of the existing law could help the reader to understand how many different and contrasting criteria are used for determining the objective arbitrability of a dispute. The Article II(1) of the NYC refers to arbitrability and is part of the NYC, which has been adopted worldwide. This article would have been the right place to provide some guidance on how to determine the arbitrability. However it does not indicate any criteria. Conversely Article II(1) of the NYC needs to be integrated by the application of a conflict of law rule to determine under which specific provision the issue of objective arbitrability should be solved. Hence, some of the existing conflict of law rules applicable to the arbitrability issue are: • Article V(2) of the NYC, which refers to lex fori as criteria to determine the arbitrability during the phase of review for the recognition or enforcement of a foreign award;55 • Article III (2)(a) of the Geneva Convention refers to the law applicable to the arbitration agreement; • Article 15(1) and 34(2)(b)(i) of the UNCITRAL Model law, propose as solution the law of the seat of arbitration;56 55 56

Arfazadeh [2001], p. 76 Hanotiau [1996], p. 391.

57

2. ARBITRAL TRIBUNALS’ JURISDICTIONAL ISSUES IN THE APPLICATION OF EU COMPETITION LAW

• A relevant solution adopted by the French and US case-law refers to the transnational concept of arbitrability.57 Under this case-law the arbitration clause should be held valid except for cases of fraud, duress, or violation of an internationally recognised concept of public policy. Critics to the above existing legal rules argue that none of the above solutions is satisfactory. The existing conflict of law rules compete among themselves causing uncertainty. Under each one of the above choice of law rules, private parties could strategically use forum shopping or choice of law clauses to avoid or to make the private enforcement in arbitration more difficult. For instance, under each choice of law rule, the party with stronger bargaining power could negotiate a distribution contract linked to the Belgian legal system. Since the Belgian law denies the objective arbitrability for distribution agreements, the application of these rules would allow the parties to avoid antitrust arbitration. The business parties would be deprived from a useful private enforcement tool and forced to use ordinary litigation. As described in the introductory Chapter this choice may result to be more costly for the private parties and dissuade them from enforcing their competition law rights.

2.1.2

The Subjective Arbitrability of Antitrust Disputes

In order to ascertain that the arbitral tribunal has jurisdiction over a specific matter of the dispute, the objective arbitrability is not enough. In the introduction of this Chapter we distinguished the objective arbitrability from the subjective arbitrability. The former, as described in the previous Section, defines whether, under a specific legal system, arbitrators have jurisdiction over antitrust issues as the matter of the parties’ dispute. Conversely, the latter defines whether in practise the arbitral tribunal has jurisdiction over the specific case at stake in light of the wording and the interpretation of the scope of the arbitration clause. 57

Namely, in an international context it is to be referred to substantive rules adapted for the needs of the international trade and there is no need to refer to any national law. See for France Cassation Civile, 20 Decembre 1993, Comit populaire de la minicipalit de Khoms El Megreb c. Dalico Contractors, JDI (Clunet) 121 [1994], 432; Cassation Civile, 7 June 2006 Comproprit maritime Jules Verne v. ABS., JDI (Clunet) [2006], 133; for the US, Eastern District of New York, 760 F. Supp. 1991, 1036.

58

Since arbitration exclusively originates from the parties’ will and has a purely contractual nature, the scope of the arbitration clause drafted and agreed upon by the parties need to cover the specific matter at stake in the dispute. If it is not proved that the parties intention was to submit such issue to arbitration, the arbitral tribunal is not entitled to address and solve the case. An arbitration clause legitimately delegates to an arbitral tribunal the power to decide a competition law dispute if it satisfies two main requirements: • firstly, the arbitration clause must be valid. It shall not only satisfy the ordinary validity requirements but it shall also not violate competition law by itself.58 Although the arbitrability and the conformity of the arbitration clause with competition law are two different and independent concepts, an arbitration clause which per se violates competition law provisions is unenforceable and void. In this case the parties did not legitimately delegate to the arbitral tribunal the powers to decide their disputes. Therefore, even leaving aside the issues of objective arbitrability, the arbitral tribunal has to verify at the preliminary stage that the arbitration clause is validly constituted, to ascertain whether it has jurisdiction over the case; • secondly, it must be ascertained whether the parties actually wanted to solve through arbitration their future disputes based on the application of antitrust law. Namely whether the scope of the validly constituted arbitration clause is wide enough to cover also competition law matters. How the arbitration clause is drafted deeply affects whether these two requirements are satisfied or not. When the parties draft an arbitration clause it can be shaped either in a pro-antitrust enforcement fashion or in such a way to make it practically difficult or virtually impossible to enforce competition law based rights. The analysis of whether the arbitration clause increases or decreases the arbitration private enforcement potential takes two different routes on the basis of whether competition law is explicitly mentioned in the clause or not. For instance, most certainly, if the parties have not mentioned competition law in 58

By “ordinary validity requirements” we mean the general principles of validity included in many legal systems. For instance the fact that a contractual clause should not be the result of duress, should not have been erroneously signed, so on and so forth.

59

2. ARBITRAL TRIBUNALS’ JURISDICTIONAL ISSUES IN THE APPLICATION OF EU COMPETITION LAW

the arbitration clause, the second requirement could be an issue. Whereas, when the parties have explicitly indicated whether competition law disputes should or should not be arbitrated, the second requirement, about the scope of the arbitration clause, is not an issue. Paradoxically regarding the first requirement, as we will discuss below, it can be more problematic when the parties have drafted an arbitration clause that explicitly provides for competition law based disputes.

2.1.2.1

The Anticompetitive Arbitration Clause

In light of the principle of party autonomy, the content of the final agreement will be determined by the parties negotiations. However, if a party has enough market power, that party is not only more likely to violate competition law, but it is also more likely to have an higher bargaining power. The latter will allow the party to negotiate the contractual clauses in its favour, including the arbitration clause. In principle this is not problematic, the vast majority of business agreements are signed between parties with different bargaining powers. The contracts can result either from hard neck-to-neck negotiations, or they can almost unilaterally be dictated by one party. In the former scenario, some clauses will favour one party, while others will favour the other party. In the latter scenario, it is more likely that most of the contractual clauses will be drafted favouring the stronger party. This does not mean that the contract will not be economically profitable also for the weaker counterparty, that can still choose whether to sign the contract and agree on its terms or not. The weaker party may also accept disadvantageous clauses as long as it believes that in toto the contract will be beneficial. Nevertheless the weaker party cannot be validly forced to renounce to its antitrust right by the party with stronger bargaining power. The legal system cannot enforce similar contracts and will deem void any such clause. Neither should it be possible to use arbitration as a tool to circumvent this obstacle. Otherwise, looking at this scenario from the antitrust arbitration viewpoint, the party that can profit from a violation of competition law has the incentive and the interest to impose, on the weaker party, an arbitration clause which is

60

drafted in such a way to impede or at least to obstruct the private enforcement of competition law.59 To define under what circumstances this behaviour can be considered as contrary to the EU competition law is not an easy task. This Section focuses on elements that need to be analysed to qualify the parties attempt to limit the application of competition law through an accurate draft of the contractual arbitration clause as illegitimate. It is worth to remember that in case an arbitration clause is in breach of the EU competition law the latter is void and it cannot validly legitimate international arbitrators to decide the case. According to EU competition law, in order to be anticompetitive, a contractual clause must have an anticompetitive effect or object.60 Moreover, according to the CJEU case-law Consten and Grundig v. Commission, first it must be determined whether the object is anticompetitive (per se infringements), then the effects produced by a clause will fall under scrutiny (to be judged with respect to economic context according to the CJEU’s case-law L’Oreal v. De Nieuwe AMCK ).61 Also in the US a similar test is used. It is derived from the joint notions of a rule of reason and per se illegality. Hence, starting from the per se violation of competition law, an arbitration clause can be prima facie economically justified in analogy to any forum-selection clause.62 As a matter of fact, even the EU Commission now requires the inclusion of arbitration clause in merger control remedies.63 Therefore similar legitimate 59

Blanke and Landolt [2010], pp. 69-88. According to Landolt an arbitration clause can breach competition law only in cases of non-unilateral conduct (Art. 101 TFEU). He claims that it is difficult to see how arbitration clauses might raise abuse of dominant position and monopolization problem.. However the scenario just described can be considered as a typical instance of market power abuse under Art. 102 TFEU. 60 Article 101 TFEU. 61 ´ Case C-56/64 and C-58/64, Etablissements Consten S.´ a.R.L. and Grundig-Verkaufs-GmbH v Commission of the European Economic Community [1966] ECR 299; case C-31/80, L’Oreal v. De Nieuwe AMCK [1980] ECR 3775. 62 The Bremen v. Zapata Off-Shore Co (1972) 407 US 1, at 13-14, ‘The elimination of all such uncertainties [upon the forum of a potential dispute resolution] by agreeing in advance on a forum acceptable to both parties is an indispensable element in international trade, commerce, and contracting’ 63 Radicati di Brozolo [2008].

61

2. ARBITRAL TRIBUNALS’ JURISDICTIONAL ISSUES IN THE APPLICATION OF EU COMPETITION LAW

practises make it difficult to argue that arbitration clauses covering antitrust disputes are commercially illegitimate or that they constitute in general a per se violation. Shifting our attention on the anticompetitive effects, Landolt clarifies that there are two specific cases which deserve specific consideration. The first one (i) when an arbitration clause limits its scope exclusively to competition law matters, and the opposite one (ii) when only competition law is excluded from the scope of an arbitration clause. In light of the general economic assumption that the one-stop principle in litigation is in the parties’ best interest, the inclusion of additional justifications for such a peculiar draft may be appropriate. For instance, in case (ii), excluding competition law from other arbitrable contractual matters could be justified in light of the less uncertain proceedings provided by national courts private enforcement. Nonetheless, as argued in the Preface, there are many inefficiencies in national courts private enforcement that arbitration can overcome, making the latter reasonable for certain international transactions. Therefore, also under case (i), it would not be easy for the parties to justify why only competition law matters have been dedicated to arbitration. There are two main arguments that could be used in favour or against the anticompetitive effects of similar arbitration clauses. On the one hand, antitrust disputes entail complex economic determinations that are quite expensive to prove in front of arbitrators (who do not have the extended fact-finding powers of national judges). For instance they involve market definitions which in some circumstances are difficult to determine without the help of a national antitrust authority.64 For instance this may be the case when private parties or arbitrators cannot easily obtain information on the market structure.65 Hence, it could be reasonable to exclude competition law from other arbitrable contractual matters. On the other hand, arbitration has some characteristics which can make it a more efficient forum. Arbitrators can be chosen by the parties among experts in the field of 64

Landolt [2006], p. 93. Nevertheless, there is no reason to assume, in general, that competition authorities are better suited to perform a market definitions than arbitral tribunals. Conversely, arbitrators may be able to provide a market definition that better reflects business realities. 65

62

antitrust law who also have a deep understanding of the parties’ commercial and business practises. Moreover, thanks to the New York Convention, in theory, arbitral award can be rendered in neutral jurisdictions, for the parties, and are more easily enforceable worldwide in comparison to national judgements.66 Therefore this argument supports an arbitration clause that limits its scope exclusively to competition law matters. However, the arguments above could be reversed and used against arbitration clause especially if they are associated to some other elements that increase the chances that an arbitration clause is considered anticompetitive. A list of these additional elements can clarify how an arbitration clause may be drafted by the parties increasing the risk that it could be challenged as producing anticompetitive effects: • Since the legal system of the arbitral seat is the one relevant during the phase of review for the annulment of the award, the parties could choose the seat of the arbitration in a jurisdiction that can be defined as unfriendly with respect to the private enforcement of antitrust law in arbitration; • The place and rules of arbitration could also affect the rules on confidentiality in arbitration. In general, arbitral hearings are private, in the sense that no one except the arbitrators, the parties, their attorneys, and witnesses may attend. Several arbitral rules also expressly provide for confidentiality (e.g. Art. 30 of the LCIA Rules, and Art. 34 of the SIAC Rules.) As Landolt says ‘the greater the confidentiality, the less effective the competition enforcement. This is because confidentiality can act as a barrier to the involvement of competition law authorities as amici or even as simple information providers.’ However, Landolt also notices that ‘Confidentiality [clauses] generally entail an exception, express or tacit, for the divulgation of information for the purpose of a legal defence.’ • the arbitration procedural rules chosen in order to accelerate the proceedings could provide the arbitrators with limited fact-finding powers. They 66

We use the expression “in theory” because as the next Chapter clarifies the enforceability of antitrust arbitral award is subject to an high degree of uncertainty.

63

2. ARBITRAL TRIBUNALS’ JURISDICTIONAL ISSUES IN THE APPLICATION OF EU COMPETITION LAW

could set a too high burden and standard of proof, that if associated with fast-track procedures is unsuitable for the assessment of an antitrust case. Landolt provides the example of the WIPO expedite arbitration rules. According to Article 10 and 12 of these rules: the deadline for submissions is 20 days; there is usually one pre-hearing exchange of written pleadings; under Article 56(a) the facts-finding stage should last not more than 3 months (whenever reasonably possible) and the award should be conveyed within a month;67 • In case the US Antitrust law is applicable the parties could choose to exclude punitive damages. As argued in the American Safety doctrine this would dilute the deterrent effect of antitrust arbitration; • Also the number of arbitrators may deeply influence the outcome of the proceeding; in case of a sole arbitrators it is difficult for the parties ex post to find an expert on all issues involved in the dispute: arbitration law, procedural law, competition law, the parties business practise, so on and so forth. All the above elements have to be taken into consideration to determine whether the arbitration clause is valid or it violates antitrust law. Again, as recurrently underlined throughout this Chapter, the outcome of this assessment largely depends on the identity of the adjudicator as well as on the choice of competition law applicable to the matter. Landolt states about this problem: ‘A court of a state signatory to the NYC would be entitled to treat an arbitration clause that is in violation of competition law as null and void, inoperative or incapable of being performed. But there is a question as to the applicability of competition law from a state that is not that whose law governs the arbitration clause. Since the seat of an arbitration tends to be chosen for its neutrality, the chances are that the courts of the seat will prove to be unconcerned as to a violation of competition law of such a third state, that is neither that of the seat nor that of the lex cause. A court requested to enforce an arbitration award is also entitled to refuse enforcement of the arbitral award if they find that the 67

Blanke and Landolt [2010], pg 74

64

arbitration clause was invalid under the law to which it was subject by the parties, which rarely occurs, or under the law of the seat of arbitration. Again, it is doubtful that the competition law of the seat of the arbitration will be engaged, and therefore a violation of competition law of a third state would not imperil the arbitration award on this basis, although it may well be on another of the few bases designated under the NYC, that is, as a violation of public policy.’68 Nevertheless, it is worth noting that even an invalid arbitration clause could be saved. Also this aspect depends on the identity of the adjudicator and on the law applicable to determine the validity of the arbitration clause. In some jurisdictions the adjudicator will never interfere with the parties autonomy, while in others this interference is considered acceptable. For instance, even if considered anticompetitive, the provisions listed above could be severed from the arbitration clause, as long as this procedure does not violate the very rationale behind parties’ choice of arbitration instead of traditional litigation. An example of this violation of rationale could be the case when the parties choose expedited arbitration procedures. The parties could consider such provision as the most important, if not the only reason for choosing arbitration. Without the expedite procedures the arbitration clause could lose its intent and its rationale in the context of the contract. Therefore, some adjudicators and some legal systems, as the US one, look suspiciously and criticise the adjudicators’ intrusion in the definition of the parties’ will. Conversely in the European context, adjudicators adopt a more paternalistic attitude and tend to amend the parties’ will wrongfully expressed in contracts .69 To conclude, an arbitration clause explicitly addressing competition law disputes can be drafted in such a way to impair the private enforcement of competition law through arbitration. In any case the relevance of this problem should not be overestimated. In practise, it can be difficult to limit the enforcement of competition law in arbitration while ensuring that the arbitral tribunal has the required set of legal tools necessary to properly apply other relevant law in the 68 69

Blanke and Landolt [2010], p. 78. Blanke and Landolt [2010], pp. 69-88

65

2. ARBITRAL TRIBUNALS’ JURISDICTIONAL ISSUES IN THE APPLICATION OF EU COMPETITION LAW

arbitration proceedings. Such clauses are incompatible with arbitration itself and would not be considered anticompetitive, but they will simply be declared void, or not validly constituted arbitration clauses.

2.1.2.2

Is Private Enforcement Within the Scope of a General Arbitration Clause?

The previous Section deals with the validity of an arbitration clause that explicitly includes or excludes competition law disputes from its scope. The same set of problems is faced also by arbitration clauses which do not mention the matter of antitrust disputes. However, unlike the former case, in the latter case, when the scope of the arbitration clause is generic, it must be ascertained whether the arbitral tribunal has jurisdiction to solve disputes based on the application of competition law. For example a general arbitration clause could read similar to this: ‘All disputes arising out of, or in connection with, the present contract shall be finally settled under the rules of arbitration of [the name of an arbitration institution], by one or more arbitrators appointed in accordance with the said rules.’ The question is whether or not the parties intended to include also the private enforcement of competition law within the wording ‘all disputes arising out of or in connection to the present contract.’ A positive answer to this question means that disputes based on competition law should be treated as disputes based on any other legal provision and be automatically included within the scope of a general and broadly formulated arbitration clause. Conversely it can be argued that, due to the specificity of this mandatory field of law, the arbitration clause should explicitly mention antitrust disputes. In other words, the question is whether the inclusion of antitrust disputes in the scope of general arbitration clauses should be accepted or not. It is clear that any arbitration clause will be subject to the adjudicators’ interpretation to ascertain that antitrust private enforcement falls within its scope.

66

The reason is that it has to be proven that the private parties wanted to delegate the solution of this disputes to private arbitrators. Interpretation of the contractual clauses is a legal technique which depends on and is guided by the applicable law or more specifically by the rules of contractual construction under the applicable law. For this reason also this issue falls under our interest and is of relevance for the present analysis. There are opposite and contrasting views upon what should and what should not be considered as covered by a general and common standard arbitration clause. As argued by Landolt, several characteristics of competition law allow practitioners to argue for and against the inclusion of competition law within the scope of such clauses:70 • Competition law based rights do not have a contractual nature and are not the result of a contractual interaction between the parties. On the contrary they have a non-contractual nature, in fact most countries classify antitrust law within the field of administrative law. Hence, it could be argued that the automatic inclusion of antitrust based disputes in general arbitration clauses is not consistent with the wording of a clause that refers to disputes based on the contractual relationship.71 The easiest counterargument points out that, since competition law has the nature of mandatory rules, it is automatically part of the contractual relationship and does not need any explicit reference to be applicable. • Competition law based actions for damages can have a tortious nature and they may be considered as not included within the scope of a contractual arbitration clause. Under article 101(2) TFEU, anticompetitive clauses are automatically void and null. According to the wording of the CJEU in Courage v. Crehan anticompetitive behaviour creates on the counterparty 70

Blanke and Landolt [2010], pp. 69-88 Blanke and Landolt [2010], p. 80. As Landolt points out an argument of this sort was made in England case, ET Plus SA v. Welter [2005] EWHC 2115 (Comm) at para. 38, by Gross J.: ‘For the Claimants, Mr. Englehart submited that... [t]he claims advanced in the Claim Form and PoC by ET Plus against Eurotunnel were not based in any way on the contract. Clause 24 could not be stretched to cover the tortious claims advanced here.’ 71

67

2. ARBITRAL TRIBUNALS’ JURISDICTIONAL ISSUES IN THE APPLICATION OF EU COMPETITION LAW

the right to request the compensation of damages. Under certain legal systems such actions for damages can be considered as contractual in nature.72 Alternatively in other legal systems damages actions are considered to be the result of an illicit harm, for instance unjust enrichment, which has a tortious nature.73 However this view is contrasting with the idea that once the parties agree to arbitrate all future disputes based on the contract they are implicitly including all issues arising out of the performance of the contract even if they are of tortious nature.74 Obviously not all tort based claims will fall within the material scope of the arbitration clause. Several tests could be applied to decide this matter. One requirement for the inclusion of such tort law based claims in arbitration is their strong connection with the performance of the contract.75 Another test requires firstly to verify how broad or narrow the scope of an arbitration clause is. Furthermore in case it is broad, it is also required to ascertain whether “collateral matters” like antitrust damage actions are included in it.76 A further test is necessary to identify the connection between the matters. This test is called the without which not test and starts from the question whether without the contract there would have been any competition law claim. A possible critic expressed by Landolt is that this standard would be too broad, including almost all competition law matters concerning the two parties of the contract. What Landolt derives from the case-law on this topic is that virtually every competition law matter has to be considered as implicitly 72

Basedow [2007], p. 112. Landolt [2006], pp. 339-340, referring to several cases: UK Garden Cottage Foods v. Milk Marketing Board [1984] 1 AC 130 (UKHL); also in Germany about Art 101, BGH, BMWImporte, WuW/E BGH 1643; BGH Cartier-Uhren, WuW/E BGH 2541; OLG Dusseldorf MetroCartier, WuW/E OLG 4407 and about article 102 OLG Stuttgart, WuW/E OLG 2018; OLG Dusseldorf Gleisanschluss, WuW/E OLG 2325. 74 In light also of Law and Economics argument in favour of the one-stop principle in litigation, Switzerland ATF 116 Ia 56 at c. 3(b), ATF 129 III 675; UK: Premium Nafta Products Limited et al. v. Fili Shipping Company Limited [2007] UKHL 40, at para 13. 75 See English case ET Plus v. Welter; also US case of JLM Industries, Inc. v. Stolt-Nielsen SA, 387 F3d 163 (2d Cir. 2004). 76 Louis Dreyfus Negoce S.A. v. Blystad Shipping and Trading Inc., 252 F3d 218, 224 (2d Cir. 2001) at par. 38; where collateral matters are defined as those ‘implicat[ing] issues of contract construction or the parties’ rights and obligations under it.’ Moreover quoting Oldroyd, 134 F.3d at 77: ‘[i]f the allegations underlying the claims touch matters covered by the parties’ contract, then those claims must be arbitrated, whatever the legal labels attached to them’. 73

68

falling within the material scope of the general arbitration clause, of course except when the parties have explicitly agreed otherwise. It is possible to summarise these viewpoints within two legal rules. On the one hand the opt-out rule: private enforcement is included in general arbitration clauses if not otherwise stated. On the other hand the opt-in rule: antitrust needs to be explicitly mentioned otherwise it will not be considered within the scope of the arbitration clause. Recent developments of national rules on international commercial arbitration show an increasingly different approach to this issue of jurisdiction. Recent reforms and recent national case-law contrast with the growing uniformity of procedures and standards developed by international arbitrators.77 The US Supreme Court, in some relatively recent decisions, adopted a new approach on the matter of authority repartition between arbitrators and courts.78 These cases are only marginally relevant for international antitrust arbitration. They are all domestic decisions on labour and employment disputes and did not affect the US policy of wide enforceability of international arbitration clauses set by the Mitsubishi case. However the discussion about domestic cases becomes interesting since in such circumstances as clarified by the case Granite Rock Co. v. Int. Broth. of Teamsters, the permissive policies favouring arbitration only apply once the validity and the existence of the arbitration agreement have been properly asserted.79 In other words, the US pro-arbitration policy works only after the existence, validity, or enforceability of the arbitral agreement is ascertained.80 In the case Rent-aCenter, the US Supreme Court declared that, in order for the arbitrators to have 77

Hanotiau [2011], pp. 89, 91; Radicati di Brozolo [2011], 663 stating that ‘The consequence of this change in State attitudes is that international commercial arbitration is less and less influenced and limited by domestic legal systems and controls’. 78 Stolt-Nielsen S.A. v. AnimalFeeds International Corp., 130 S. Ct. 1758 (2010); Rent-ACenter, W., Inc. v. Jackson, 130 S. Ct. 2772 (2010); Granite Rock v. Intern. Broth. of Teamsters, 130 S. Ct. 2847 (2010); ATandT Mobility LLC v. Concepcion, 131 S. Ct. 1740 (2011). 79 Granite Rock Co. v. Int. Broth. of Teamsters, 130 S. Ct. 2847, 2957-58 (2010). 80 Also referred to as gateway or threshold issues in Rent-A-Center, 130 S. Ct. at 2777, observing that parties can agree to arbitrate gateway questions of arbitrability, such as whether the parties have agreed to arbitrate or whether their agreement covers a particular controversy, (citing Green Tree Financial Corp. v. Bazzle, 539 U.S. 444, 452 (2003)).

69

2. ARBITRAL TRIBUNALS’ JURISDICTIONAL ISSUES IN THE APPLICATION OF EU COMPETITION LAW

jurisdiction over the issue of validity, enforceability and scope of an arbitration clause, the parties need to explicitly mention it in the arbitration clause. Otherwise, if the parties have not expressly provided on this point, the jurisdiction belong to the national courts.81 As noted by Naon, in this case the US Supreme Court applied the principle of separability of an arbitration clause in a rationale similar to the one embodied in the Prima Paint and Buckeye cases, which do not associate the principle of separability with the principle of kompetenz-kompetenz with respect to matters regarding the validity of the arbitration clause.82 As mentioned in the introductory Chapter, the doctrine of separability provides that an arbitration clause is autonomous, hence separable, from the main contract to which it belongs. Therefore under this principle the arbitration clause is not affected by the validity of the contract.83 In principle, under a wide definition of the doctrine of kompetenz-kompetenz challenges to the validity of the contract or the jurisdiction of arbitrators, namely over the validity of the arbitration clause, are to be determined by the arbitrators. Conversely, in the cases Prima Paint and Buckeye the US Supreme Court seems to adopt a narrow definition of the doctrine of kompetenz-kompetenz. In fact, it stated that challenges to the validity of the arbitration clause itself have to be determined by the national courts. In 2010, the US Supreme Court confirmed again the principle that the national courts have to decide upon issues regarding the arbitral tribunal’s jurisdiction, unless explicitly otherwise provided by the parties in the arbitration clause.84 Notwithstanding this restricted application of the principle of kompetenz-kompetenz the US legal system is still an arbitration friendly jurisdiction. In fact the US policy in favour of arbitration is mostly untouched, since, in practise, the courts solve any ambiguity concerning the scope of the parties’ arbitration clause in favour of the subjective arbitrability.85 Nevertheless, this determines a special path for solving issues on jurisdiction. 81

Rent-A-Center, W., Inc. v. Jackson, 130 S. Ct. 2778 (2010). Prima Pain Corp. V. Flood and Conklin Mfg. Co., 388 U.S. 395 (1967); Buckeye Check Cashing, Inc. v. Cardegna, 546 U.S. 440 (2006). 83 Smit [2002]. 84 Granite Rock Co. v. Int’l Broth. of Teamsters, 130 S. Ct. 2847, 2858 (2010). 85 Granite Rock Co. v. Int Broth. of Teamsters, 546 F.3d 1169, 1177 (9th Cir. 2008). 82

70

To ascertain the arbitrability of competition law disputes, US national courts will retain jurisdiction, unless arbitration clauses explicitly entitled the arbitral tribunal to deal with this issue. In other words, the rule applied in the US is the opt-in rule. This rule is the exact opposite of the one generally adopted in Europe. For instance in France both the doctrine of separability and of kompetenz-kompetenz are largely applied in such a way to entitle both domestic and international arbitrators to address any issues regarding their own jurisdictions without the need of explicitly opting-in.86 A national court, in front of which a dispute about an arbitration agreement is brought, has to decline its jurisdiction and send the parties to the arbitral tribunal. The only limitation to this rule is in case of an arbitration agreement manifestly void or inapplicable. This needs to be assessed by the national court only on a prima facie inquiry.87 Therefore, contrary to the US solution, in Europe, in light of the one-stop adjudication principle, questions upon the validity, existence or scope of the arbitration clause do not need to be explicitly mentioned within the arbitration clause to entitle the arbitral tribunal with the matter. The effect of such inconsistency among legal systems results in the application of different procedures and different legal rules depending on which court faces the second-look review of the award. As clarified in the introduction of this Chapter this inconsistency in antitrust arbitration produces uncertainties and increased costs for the parties. Therefore, also the issue of determining the scope of the arbitration clause is strictly affected by both the question of choice of law applicable to the dispute and the question of who is the adjudicator entitled to decide on the issue of arbitrability. The Law and Economics methodology adopted in the last Section of this work should help to shed some light on incentives and trade-offs at stake when applying 86

See Article 1448 of the French Code of Civil Procedure; and also the Decree 2011-48 dated 13th of January 2011, on the Reform of Arbitration. 87 Loquin [2011]; Article 1448 of the New French Code of Civil Procedure.

71

2. ARBITRAL TRIBUNALS’ JURISDICTIONAL ISSUES IN THE APPLICATION OF EU COMPETITION LAW

each one of these rules.

2.2

Antitrust Arbitration Needs a Uniform Choice of Law Rule

The previous Section clarified what it means to determine whether the dispute is arbitrable, namely whether international arbitrators have jurisdiction over the antitrust matters at stake in a dispute. Once this is ascertained, the following step for the arbitrators will be to identify the law applicable both to the merits of the case and to the arbitral proceeding itself. The choice of law issue, that is the subject matter of this Section, is typically a preliminary matter, although it is strictly connected to the merits of the case. Hence it can be considered a bridge topic between the jurisdictional part of this work and the part dealing with questions of merits.

2.2.1

Conflict of Law and Extraterritorial Application of Competition Law

An example can help us introduce better the topic of this Section. This simplified example, although fictional, is surprisingly close to reality. Coca-Cola is an international company, head-quartered in the US, that has a dominant position in the market of sparkling beverages. The company sells its products worldwide through local distributors, including European undertakings, with whom it signs vertical distribution agreements. Let us assume that Coca-Cola has the bargaining power to draft such distribution agreements including vertical restraints that can potentially restrict competition. For instance, we can think of a specific rebates structure linked to minimum purchase quantities, supply of refrigerators linked to their exclusive use for Coca-Cola beverages, and other similar provisions.88 This scenario revolves around issues on the validity of vertical agreements under 88

This example comes from a real case which was settled between Coca-Cola and the EU Commission. Coca-Cola proposed remedies accepted by the EU Commission, who considered them capable to guarantee the compliance of the distribution agreements with the EU competition law.

72

different antitrust regimes.

2.2.1.1

The Conflict of Law Issue in Antitrust Arbitration

The Commission has taken a stricter position towards such vertical restraints in comparison to the US antitrust agencies. Two reasons have been proposed to justify the different approaches taken towards vertical restraints: the first reason is the strong influence that the economic model proposed by the Chicago School has exerted upon the US agencies and courts, and the second reason is the European Union’s objective of achieving an integrated market and preventing the segmentation of this market into different national markets.89 This means that selective distribution strategies that have no trouble to pass the US’s rule of reason test, may not only violate Article 101(1) TFEU but may prove ineligible for individual exemption ex Article 101(3) TFEU. This is particularly true in cases of indirect minimum resale price maintenance, some territorial and customer restrictions, restrictions to sell only to end-users imposed on retailers in a selective distribution system, restrictions on cross supplies within a selective distribution system, and restrictions on component suppliers to sell the components they produce to independent repairers or service providers, and attempts to add export bans to agreements limiting a distributor to one Member State territory. Similar kinds of vertical agreements are common also to other business sectors or types of products. They are quite relevant for the world economy, since they may have the value of millions of Euros and produce effects on several national markets as well as on millions of final consumers. As in the Coca-Cola example, these agreements may be in breach of the EU competition law while being completely legal according to the US antitrust law. Therefore, such contracts between US and EU based undertakings are the cornerstone of this work. In fact, they may also contain an arbitration clause. This means that if in our Coca-Cola example the EU Commission condemns Coca-Cola for anticompetitive behaviour, the arbitration clauses may be triggered by the EU distributors seeking compensation for damages due to the breach of EU Competition law, within follow-on disputes in arbitration. The solution that arbitrators provide to the conflict of law issue 89

Cooper et al. [2004]; Zekos [2008b].

73

2. ARBITRAL TRIBUNALS’ JURISDICTIONAL ISSUES IN THE APPLICATION OF EU COMPETITION LAW

becomes the key element to determine the outcome of the dispute. 2.2.1.2

The Extraterritorial Application of Competition Law

In the context of a cross-border business transaction, like the above, more than one competition law regime may be applicable to the dispute. The reason for this is based on the extraterritorial effect of national antitrust rules. The extraterritorial application of competition law is the doctrine by which one law enforcement entity can seek to enforce its law outside of its territory under a conflict of law provision.90 With respect to the EU law, the EU antitrust sanctions can be imposed for conducts that offend EU competition law perpetrated by undertakers outside of EU.91 The Court of Justice of the European Union has, also, supported the Commission’s view that the EU competition law has extraterritorial effect.92 However, here we are not concerned with the public enforcement issues, on the contrary we focus here on the private enforcement side. To summarise, international agreements resulting into disputes with, or exclusively between, non-EU parties performed in one jurisdiction but producing 90

Hood [1982]. In this work we use the expression conflict of law interchangeably with the expression private international law. 91 Commission notice, Guidelines on the effect on trade concept contained in Articles 81 and 82 of the Treaty [Official Journal C 101 of 27.4.2004], Par. 100 reads that: ‘Articles 81 and 82 apply to agreements and practices that are capable of affecting trade between Member States even if one or more of the parties are located outside the Community(78). Articles 81 and 82 apply irrespective of where the undertakings are located or where the agreement has been concluded, provided that the agreement or practice is either implemented inside the Community(79), or produce effects inside the Community(80). Articles 81 and 82 may also apply to agreements and practices that cover third countries, provided that they are capable of affecting trade between Member States. The general principle set out in Section 2 above according to which the agreement or practice must be capable of having an appreciable influence, direct or indirect, actual or potential, on the pattern of trade between Member States, also applies in the case of agreements and abuses which involve undertakings located in third countries or which relate to imports or exports with third countries.’ 92 Case 48/69, Imperial Chemical Industries Ltd v. Commission (Dyestuffs) [1972] ECR 619, which starts the single economic entity doctrine and considers three non-EU undertakings liable for price-fixing with effects in EU market through their subsidiaries; see also case 89/85, Ahlstrm Osakeyhti v Commission (Wood Pulp) [1988] ECR 5193 para 11-23, and Commission Decision OJ 2001 L152/24 para 182, where the implementation of an illegal agreement within the EU has been used to justify the application of EU competition law to non EU undertakings.

74

effects in the market of another jurisdiction, will typically imply a conflict of law issue. In other words, the adjudicator has to determine which of the involved overlapping competition law disciplines should be applied to solve the case. What makes this issue especially important is the fact that the competition law regime that results from the solution of the conflict of law question will substantially determine whether or not there is a breach of competition law, namely it will determine who wins the dispute. For instance, in the example provided at the beginning of this Section, if the arbitrators apply the US antitrust law there will be no breach of law, while if the law applicable is the EU competition law the conduct could be considered anticompetitive.

2.2.2

Choice of Law Rules

As we have mentioned above, because of its extraterritorial effect, the application of competition law to international transactions is potentially subject to conflict of law issues. There may be several competing antitrust regimes that arbitrators and national judges may have to consider in order to solve a conflict of law issue. One may think that the antitrust law applicable will be identified according to the private international rules of choice of law followed by national courts. All the same, it has been noted that the solution is not that simple. One reason for this is based on the different approach followed by national courts and by arbitrators when dealing with this issue. There are two different approaches followed respectively by national judges and by international arbitrators on the issue of conflict of law. This is due to the fact that, as opposed to national courts, arbitral tribunals sitting in a State do not have the duty to apply the mandatory rules of the law of that State to the merits of the case.93 Arbitrators do not share with judges the same concept of forum and cannot rely on a system of conflict of law of their own. There is actually no conflict of law provision specifically designed for arbitration.94 It is even an open question whether the international conventions on the conflict of 93 94

Landolt [2007a]. Landolt [2007a].

75

2. ARBITRAL TRIBUNALS’ JURISDICTIONAL ISSUES IN THE APPLICATION OF EU COMPETITION LAW

law can be applied by arbitrators, given the fact that they are designed for courts and do not mention arbitration.95 Therefore the two approaches to the conflict of law question, one followed by arbitrators and the other by national judges are substantially different and need to be reconciled. 2.2.2.1

National Courts’ Approach to Conflict of Law Issues

National judges with jurisdiction over the case sit in a court which is located in a specific forum governed by a legal system. State judges rule their decisions on behalf of the State from which they derive their authority. Their legal system has a set of rules on private international law that guides the judges’ decision on the law applicable to the dispute.96 It is true that according to the rules of private international law, parties may be granted the option to choose ex ante, within their contract, the law applicable to the agreement. However the private international law provisions of the lex fori determine the limits of the parties freedom to choose rules that are different from the one of the forum of the judge with jurisdiction. When setting these limits the lex fori can also determine some provisions that have to be regarded as mandatory rules and that cannot be optedout by the parties, such as antitrust rules. A national judge must always apply these mandatory rules, regardless of whether a given relationship is or is not international. Thus, as far as a judge is concerned, there is always a law that must be applied, either directly or through his country’s own system of conflict of law. For instance, the US national courts’ conflict of law rule set for antitrust extraterritorial application is stated in the Foreign Trade Antitrust Improvements Act (FTAIA), Subsection 1(A). In order to apply the US antitrust law, the test of qualified effects requires that the allegedly anticompetitive conduct aims to produce or concretely produces a substantial or reasonably foreseeable direct effect on the US market. The European Court of Justice followed on the same path through the implementation theory which in practise was applied in the same way as the US test of 95 96

Landolt [2007a]. It is generally called lex fori.

76

qualified effects.97 Furthermore both jurisdictions are supposed to take into consideration the role of comity, namely mutual respect of the foreign jurisdiction, in order to limit the effects of the extraterritorial application of their respective antitrust disciplines.98 However, in practise the national courts’ approach is quite different. Given the effects of the global economy on the internal market of each jurisdiction, it has been argued that national courts, when there is a conflict of law question, simply apply their own competition law provisions.99 The reason may lie on judges’ general resistance to apply a less familiar foreign law. In any case, as mentioned above, national judges do not undertake an in-depth analysis on the conflict of law issue but simply verify whether the extraterritorial application of their mandatory rules is wide enough to cover the merit of the specific case at stake.100 Multiple empirical studies conducted on this matter have proved consistently that if the lex fori is applicable to the case, national judges tend to solve the conflict of law question in favour of its application.101 To summarise, in theory national courts follow private international rules of choice of law, but in practise, it has been argued that they do not engage in an unbiased conflict of law analysis. National judges are linked to a legal system. When competition law is involved they verify whether extraterritoriality of their antitrust law can be extended to cover the case at stake.102 97

Case 89/85, Ahlstrm Osakeyhti v Commission (Wood Pulp) [1988] ECR 5193 defines the implementation theory; subsequently the Court of First Instance described how it should practically be applied (the same was as the US test of qualified effects) in the case T-102/96 Gencor Ltd v Commission of the European Communities [1999] ECR II 00753. 98 It should be noted that the US Supreme Court limited the use of the principle of comity in its decision Insurance Co. v. California , 509 U.S. 764, del 1993. It stated that the comity, intended as balance of the conflicting states’ interests to the application of their national law, should be applied only in those exceptional circumstances when the application of the US law would substantially mean a violation of the other state’s rule of law. Conversely in Europe, both the EU Commission and the CJEU have made clear in the case-law mentioned above that they consider the EU competition law applicable in all those cases which produce a substantial anticompetitive effect, either direct or indirect, within the internal market. 99 Landi and Rogers [2007]. 100 Landi and Rogers [2007]. 101 Borchers [1992]; Solimine [1989]; Thiel [2000]. See also O’Hara et al. [2009], pp. 641-642. 102 Landi and Rogers [2007].

77

2. ARBITRAL TRIBUNALS’ JURISDICTIONAL ISSUES IN THE APPLICATION OF EU COMPETITION LAW

2.2.2.2

Arbitral Tribunals’ Approach to Conflict of Law Issues

The arbitral proceeding is not linked to a specific legal system nor to a unique choice of law rule. The award itself may be enforced in more than 160 different countries, according to the New York Convention.103 It is worth remembering that arbitrators carry out a mission given to them by the parties. The mission is to solve the dispute providing an enforceable award. Hence, for the international arbitrator, all national laws have the same value and none of them has a privileged status. Therefore, as a general principle, in arbitration the law applicable both to the merits and to the procedure is subject to parties’ choice of law either in the arbitration clause or within the contract where the latter can be found. When there is no such choice of law by parties, there is an internationally shared rule granting to arbitrators the freedom to choose the law they believe most appropriate in order to determine the norms applicable to the merits of the case.104 However, this different approach to the conflict of law taken by arbitrators, with respect to the one taken by national judges, becomes problematic in the context of rules of mandatory application such as antitrust rules. When deciding whether or not to apply a relevant set of mandatory rules, international arbitrators must be mindful of the award enforceability.105 In fact it has been rightly pointed out that international arbitration is ‘above, or outside the national legal systems until its final product (the award) lands before a national court or even several national courts within setting aside or enforcement proceedings’.106 At that point arbitrators’ freedom to determine the law applicable meets the limit of the second-look review of the award by a national court.107 Nevertheless, with regard to the application of mandatory norms, international arbitrators do not 103

The New York Convention of 10 June 1958 on the Recognition and Enforcement of Foreign Arbitral Awards. 104 See Article VII(1) of the European Convention on International Commercial Arbitration, Article 1496 of the French New Code of Civil Procedure, Article 28 of the United Nations Commission on International Trade Law (UNCITAL) Model Law, Article 187(1) of the Swiss Law of Private International Law, Article 17(1) of the ICC Rules. 105 Article 26 of the ICC Rules of Arbitration emphasises this issue by providing that the arbitrator ‘shall make every effort to make sure that the award is enforceable at law’. 106 Radicati di Brozolo [2005]. 107 The second-look review will be thoroughly discussed later in this Section.

78

have per se an overriding duty to ensure that concepts and policies embodied in national or international state law are respected. They do not have to explain that the conflict of law rule they apply is the one that is applicable according to a specific system of conflict of law. But they cannot either support their decision to apply a rule defined as mandatory in a specific system by the mere fact that this rule claims to be applied. When competition law matters are involved the arbitration general choice of law rule described above will contrast with the one of national courts. Competition law provisions are intended by states to be applicable irrespectively of parties’ choice of law.108 As clarified by the US Supreme Court in the Mitsubishi decision, it cannot be tolerated that in business contracts private parties avoid the civil consequences of an anticompetitive conduct using an arbitration clause and a choice of law provision.109 In the famous footnote 19, the Supreme Court stated that: ‘In the event of the choice-of-forum and choice-of-law clauses operated in tandem as a prospective waiver of a party’s right to pursue statutory remedies for antitrust violations, we would have little hesitation in condemning the agreement as against public policy.’110 2.2.2.3

List of Choice of Law Rules in Antitrust Arbitration

The several conflict of law rules that may come at stake when the arbitral tribunal has to solve the conflict of law issue all provide unsatisfactory solutions to this problem: • the lex contracti resulting from the choice of law clause included by the parties in the contract. Although the possibility for private parties to choose the law that should regulate their agreement is widely accepted, this choice is incompatible and cannot affect the application of mandatory rules, as defined above. Competition law is intended to be applied to a contract irrespectively from the lex contracti chosen by the parties. It can be the 108 109

Landolt [2007a]. Mitsubishi Motors v. Solar Chrysler-Plymouth, 473 U.S. 614, 105 S. Ct. 3346 (1985) par.

637. 110

This is a footnote in the Mitsubishi decision.

79

2. ARBITRAL TRIBUNALS’ JURISDICTIONAL ISSUES IN THE APPLICATION OF EU COMPETITION LAW

law of a Member State or any other law, for instance, US law or Russian law, as per our examples in the introduction of this Section. Should the parties have chosen the US law or the EU law, the international arbitrators would still have to decide which are the antitrust rules that are nevertheless applicable. This is a result of the mandatory nature of competition law provisions; • the lex arbitri, namely the law of the seat of arbitration. This is the law of the place of arbitration and its mandatory rules must be taken into consideration since that is the forum of the judge where the parties may challenge the award, requiring its second-look review by national courts. However this choice is not entirely satisfactory, since also the seat of arbitration results from the parties’ free choice when negotiating the contractual arbitration clause. All the same, arbitrators are aware that, by not applying the mandatory rules of the seat of arbitration, they put at risk the validity and enforcement of the award.111 When the seat of the arbitration is located in the EU, as lex arbitri, the EU competition law is part of the public policy within the domestic law of each Member State and the award can easily be set aside if it does not apply EU competition law; notwithstanding the fact that, in theory, the ground for the review of the award by national courts should be relevant only for procedural issues and should not affect the merit of the case. This is true especially when the award has to be enforced in another country and there is no link between the dispute and the seat of arbitration. To make things more complicated, it has also been noted that some countries are open to enforce arbitral awards which have been set aside within the country of the seat of arbitration;112 • the law of place of enforcement of the final award. The problem with this rule is the fact that often in international arbitration the place of enforcement largely depends on the winning party. Therefore this rule is unsatisfactory because it is a typical example of circular reasoning. As explained before, the winning party may largely depend upon which antitrust regime 111 112

see Art V(1)(e) NYC. Na´ on [2012].

80

arbitrators decide to apply. For instance, in our Coca-Cola example at the beginning of this Section, if the arbitrators apply the EU competition law, there will be a breach of law and the winning European party will seek enforcement against Coca-Cola in the US. In case the arbitrators decide to apply the US antitrust law the result will be the opposite; • a foreign loi de police of mandatory application. This rule is based on the close connection of a legal system with the case. However, when the case is international, it is still difficult to decide which rule should be used to determine the close connection. For instance, the green paper on damages suggests that courts should apply the law of the market where the victim of the anticompetitive conduct is situated,113 in accordance with article 5 of the Rome II Convention.114 However, this rule does not necessarily correspond to the one applied by national judges and does not solve the concerns derived from the second-look review in front of national courts; • it has been proposed also to use a rule similar to the one contained in the Convention on the Law Applicable to Contractual Obligations signed in Rome on 19 June 1980. The most celebrated instance of the application of mandatory norms is Article 7(1) of the Rome Convention which reads as follows: ‘When applying under this Convention the law of a country, effect may be given to the mandatory rules of the law of another country with which the situation has a close connection, if and in so far as under the law of the latter country, those rules must be applied whatever the law applicable to the contract. In considering whether to give effect to these mandatory rules, regard shall be given to their nature and purpose and to the consequences of their application or non-application.’ The Rome Convention has now been substituted (except for Danish courts) by the European Community Regulation Rome I (European Council Regulation no. 593/2008 of 17 June 2008 on the law applicable to contractual obligations (Rome I), OJ 2008 L177/6 of 4 Jul. 2008). Article 9(3) of the Rome I Convention is the functional equivalent of Article 7(1) of the Rome Convention, but maybe 113 114

See EU Green Paper on Damages in Private Antitrust Litigation at n.10. European Council Regulation 864/2007.

81

2. ARBITRAL TRIBUNALS’ JURISDICTIONAL ISSUES IN THE APPLICATION OF EU COMPETITION LAW

not as well suited to cover competition law mandatory rules as Article 7(1), since it provides as follows: ‘Effect may be given to the overriding mandatory provisions of the law of the country where the obligation arising out of the contract have to be or have been performed, in so far as those overriding mandatory provisions render the performance of the contract unlawful. In considering whether to give effect to those provisions, regard shall be had to their nature and purpose and to the consequences of their application or non-application.’ Therefore it is possible to argue that the wording of Article 9(3) of Rome I contemplates a narrower category of mandatory norms than does Article 7(1) of the Rome Convention. However also this rule is unsatisfactory because of its indeterminate character. In light of the above analysis only one thing is certain: arbitrators seek to take into account the parties’ legitimate expectations when exercising their freedom to determine the law applicable to the merits. Therefore in practise, the solution adopted by arbitrators is the use of a cumulative approach, choosing the antitrust regime that is identified by most of the described choice of law rules. However, while this solution may work for many cases, it does not provide a unique, objective and satisfying conflict of law rule. Phillip Landolt, one of the most prominent scholars in this field, suggests that the application of competition law by international arbitrators may be thought as the equilibrium ‘between two poles represented at the one end by the party will and at the other by the will of the State whose competition law is at issue.’115 To conclude, arbitrators aim to provide private parties with a valid and enforceable solution to their dispute. As such, the arbitral tribunal is not in charge of protecting the public policy of a determined national legal system. Arbitrators are not the guardian of the EU public policy either. However, arbitrators lack the second characteristic of a judicial power (the ius imperium), namely the power to enforce their decisions. As we have seen in the previous Section and in the introductory Chapter, the main point emerging from the existing case-law on antitrust 115

Landolt [2007a].

82

arbitration is that, even if there are public policy interests involved in antitrust arbitration, the lawmaker widely accepts the arbitrability of competition law issues on the assumption that the public policy interests will be safeguarded in the second-look review of the award by national courts. Consequently, the attitude of the courts exercising the second-look review of the final award becomes of fundamental importance for the conflict of law rule adopted by arbitrators. Before delving into the legal analysis of the second-look review of conflict of law issues, it will be useful to clarify some recurrent, and often confused, conflict of law concepts that are relevant in antitrust arbitration.

2.2.3

Conflict of Law Notions Relevant for Antitrust Arbitration

Within the New York Convention on the Recognition and Enforcement of Foreign Arbitral Awards there are two fundamental concepts that limit the circulation and enforceability of international awards among the signatory States, namely the arbitrability and the public policy exception. Article V(2) of the NYC refers to them as two separate grounds for setting aside an arbitral award. Hence, they are the two key elements that need to be taken into consideration for the recognition and enforcement of foreign arbitral awards. Moreover, they are the only way for States to control the activity of international arbitral tribunals. For this reason we started this Chapter analysing the arbitrability of competition law matters with a specific focus on two issues: (i) identifying the adjudicator entitled to assess the matter, and (ii) solving the conflict of law applicable to decide the question of arbitrability as defined above. This Section moves from the concept of antitrust arbitrability to the issues raised by competition law in connection to the concept of public policy as defined under the New York Convention.116 In practise the two elements are strictly related. However, since in some cases they have been wrongfully confused, it is important to distinguish them before starting the in-depth analysis of this topic.117 Certainly an arbitral tribunal will 116 117

Blanke and Nazzini [2008a]. See the Belgian Supreme Court case law: 28 Jun 1979 Audi-NSU v. S.A. Adelin Petit, 1

83

2. ARBITRAL TRIBUNALS’ JURISDICTIONAL ISSUES IN THE APPLICATION OF EU COMPETITION LAW

violate the national public policy of a State solving a case on a matter that is defined as not arbitrable by mandatory rules of that legal system.118 The same will hold true when a dispute is considered not arbitrable on the basis of an arbitration clause that is in itself in violation of the public policy of the state of recognition or enforcement.119 However, these two examples are substantially different from the case of a dispute to be decided by arbitrators, which revolves, in its merits, around a violation of national public policy. The fact that a national mandatory norm is at stake does not imply that the matter of the dispute is not arbitrable. Neither an arbitral award, itself in violation of national public policy, makes the dispute not arbitrable. Therefore, the two concepts are different and independent, although in some cases they can overlap. In other words, the rationale behind the public policy exception is different from the one of the arbitrability principle and can be summarised as follows: if the law applied by the arbitrators happens to violate the international public policy of the country where enforcement of the award is sought, that incompatibility will be a valid ground for refusing enforcement. In this case, however, it is not the choice made by arbitrators to apply a particular law (arbitrable or not) that is in question. On the contrary, it is the impact of the solutions of that law on the merits of the case that is the main cause for concern.120 Therefore the two concepts must be dealt with separately, while keeping in mind their correlation. This implies that the parties choice of law rule used to deal with the problem of arbitrability and the issue of the law applicable to the merit of a dispute must be kept aside. The reason why the concept of public policy is relevant for the present research is the fact that the EU competition law is not only part of the legal order of the EU Member States, it is even part of their own public policy. 121 However, Pas 1260, RCJB, 332, Note C. Elst. Here the court, imprecisely states that they are the same identical thing. 118 This is the case discussed above about lack of objective arbitrability. 119 This point is identical to the matter of lack of subjective arbitrability discussed above on the ground of an invalid arbitration clause 120 Shelkoplyas [2003], p. 171. 121 Case C-126/97, Eco Swiss China Time Ltd v Benetton International NV [1999] I-3055, see also the decision of the Court of Appeal of Paris, 18 November 2004, Thal´es AirDefence v. GIE Euromissile (2005) Rev. Arb., 751.

84

competition law provisions are not only part of the EU public policy, they are also defined by States as mandatory rules. According to Hochstrasser, ‘mandatory rules of law (loi de police in French) are defined as imperative provisions of law which must be applied to an international relationship irrespective of the law that governs that relationship; they are a matter of public policy (ordre public), and, moreover, reflect a public policy so commanding that they must be applied even if the general body of law to which they belong is not competent by application of the relevant conflicts-of-laws rule.’122 It is the imperative nature per se of such rules that makes them applicable. One is thus led to conclude that there is an ‘approach to mandatory rules of law’ different from the classical method of conflict of laws.123 Since also the concept of mandatory rules often overlaps with the different concept of public policy, it is important here to distinguish them carefully. On the one hand, international mandatory rules are rules that aim to be applied in a predetermined field in any circumstance, irrespective of the ordinary rules of conflict of law.124 They operate ex ante, intervening before any decision upon the applicable law, preventing the intervention of any choice of law rule. However, international mandatory rules can overlap, conflict and compete, therefore they must be linked to specific characteristics of the case. On the other hand, public policy intervenes ex post as a basis for refusing to give effect to rules designed as applicable by the conflict of law provisions, but that are incompatible with a legal system.125 Thus it is in their form of mandatory norms that international arbitrators are confronted with the problem of the application of conflicting competition rules. Conversely, it is on the ground of violation of the national public policy that an arbitral award is set aside by a national judge in the second-look review phase of recognition or enforcement. In light of these clarifications, we can say that international arbitrators are 122

Hochstrasser [1994]. Mayer [1986]. 124 The legal system does not allow private parties to avoid their application through a choice of law contractual clause or the choice of the arbitration seat. 125 Shelkoplyas [2003], p. 171. 123

85

2. ARBITRAL TRIBUNALS’ JURISDICTIONAL ISSUES IN THE APPLICATION OF EU COMPETITION LAW

granted the power to apply mandatory rules that belong to a legal systems’ public policy, such as antitrust rules, and to assess the civil consequences of their breach.126 However, their potential assessment of a conflict of law problem falls under some kind of control by national judges that will be in charge of the recognition and enforcement of the arbitral award.127 This control is legally based on the concept of public policy violation.

2.2.4

Analysis of the Effect of the Second-look Review on the Suitability of Arbitral Tribunals’ Choice of Law

The second-look review in different legal systems has taken a trend that is all but uniform. The extent of the review and its effects may drastically differ in connection to the legal system under which the parties try to reform, recognise or enforce the arbitral award. Different jurisdictions may have various approaches to the extent and the nature of the inquiry during the second-look review. The result is that where there is a limited review of the award the national court will not consider how the arbitrators have solved the conflict of law problem. Conversely, in other States, following the extensive review doctrine, the matter will fall under the scrutiny of the national judge, which may set aside the award or refuse to enforce it because it does not apply the antitrust law that the judge believes appropriate. Given the arbitrators’ duty to provide the parties with a valid and enforceable award, when they choose the law applicable they may systematically favour the antitrust law of the jurisdiction involved in the dispute, with the higher standard of review. The situation is even more complicated when we look at different state courts’ approach to the enforcement of an award that was set aside in the state of its seat. In the US the case-law follows strictly the wording of the NYC and does not enforce an arbitral award that was set aside in the legal system of its seat. 126

In the case of private enforcement of competition law, arbitrators decide upon the civil consequences of a violation of competition law, namely: enjoining a party to cease violating the other’s rights; awarding damages or invalidating the contract 127 McLaughlin [1995]; Case C-126/97, Eco Swiss China Time Ltd v Benetton International NV [1999] I-3055.

86

Conversely, in Europe, the French judges have recently proved that they do not consider the annulment of an award by national courts of the arbitral seat to be enough to justify the denial of the award recognition or enforcement in France.128 As mentioned in the introduction of this Section, the described cross-jurisdiction inconsistency affects also the suitability of arbitrators’ conflict of law rules and adversely influences the effectiveness of private enforcement in arbitration. One of the fundamental advantages of arbitration over traditional litigation, as dispute resolution mechanism for international transactions, is its reduced uncertainty on matters of law and less burdensome procedures. However, currently, as far as the application of competition law is at stake, this advantage appears at least diluted. 2.2.4.1

Second-look Review Doctrine in the US

Introducing a description of the Second-look review case-law in the US, we have already mentioned the famous footnote by the US Supreme Court in the Mitsubishi case, the first case-law establishing that antitrust disputes are arbitrable in the US.129 ‘In the event of the choice-of-forum and choice-of-law clauses operated in tandem as a prospective waiver of a party’s right to pursue statutory remedies for antitrust violations, we would have little hesitation in condemning the agreement as against public policy.’ This sentence leaves room for setting aside foreign arbitral awards that did not apply US antitrust law when it was extraterritorially applicable. However, this principle has been subsequently overruled by the US case-law. In Simula v. Autoliv has been stated that if international arbitrators do not apply US antitrust law, in substitution, it is also acceptable the application of a foreign competition law, as long as it does not deprive the parties from a reasonable competition law protection.130 The result is that, currently, arbitrators’ choice of antitrust law applicable to the dispute is virtually free from any challenge by national courts.131 128 129

Na´ on [2012]. Mitsubishi Motors v. Solar Chrysler-Plymouth, 473 U.S. 614, 105 S. Ct. 3346 (1985) par.

637. 130 131

Simula v. Autoliv. 175 F. 3d 716 (9th cir. 1999). Landi and Rogers [2007].

87

2. ARBITRAL TRIBUNALS’ JURISDICTIONAL ISSUES IN THE APPLICATION OF EU COMPETITION LAW

2.2.4.2

Limited and Extended Second-Look Review in Europe

On the contrary, the European approach is quite problematic. In Europe, the degree of second-look review of the award is unclear and inconsistent among Member States. The EU competition law is part of the EU public policy and is directly applicable in all Member States.132 National courts are obliged to guarantee that the EU antitrust provisions are enforced. The CJEU’s case-law133 has made clear that arbitrators are required to apply the EU competition law, given its fundamental importance for the European legal system and the functioning of the internal market.134 The CJEU also stated that national courts’ review of the award ‘may be more or less extensive depending on the circumstances’.135 However no guidance was given about the extent or the nature of the review except that: ‘it is in the interest of efficient arbitration proceedings that review of arbitration award should be limited in scope and that annulment of or refusal to recognise an award should be possible only in exceptional circumstances’.136 Conversely, Advocate General Saggio (henceforth AG Saggio), in his opinion on the same case, argued for an extensive review of the award: ‘courts which may be called upon to determine whether arbitration awards are compatible with rules of law must be allowed to carry out an effective review of the award in question.’137 This inconclusive decision created different standard of review throughout the EU national courts and jeopardised arbitrators and parties reliance on the degree of award enforceability in the EU. The different extent of the second-look review that can be found within the case-law of different national courts varies throughout a whole range running between two extremes. On the one hand, there is the 132

Case C-453/99, Courage v. Bernard Crehan [2001] ECR I-6297. Case C-102/81, Nordsee Deutsche Hochseefischerei GmbH v Reederei Mond Hochseefischerei Nordstern AG & Co KG [1982] ECR 1095; case C-393/92 Gemeente Almelo et al. v. NV Energiebedrijf IJsselmij [1994] ECR I-1277; and case C-126/97, Eco Swiss China Time Ltd v Benetton International NV [1999] I-3055. 134 This point will be thoroughly analysed below in the Chapter on the Arbitrators’ duties to apply EU competition law ex officio. 135 Case C-126/97, Eco Swiss China Time Ltd v Benetton International NV [1999] I-3055, par. 32. 136 Case C-126/97, Eco Swiss China Time Ltd v Benetton International NV [1999] I-3055, par. 48. 137 Opinion of Mr Advocate General Saggio delivered on 25 February 1999 on China Time Ltd v Benetton International NV, at par. 32. 133

88

limited review approach which is satisfied by a prima facie control about whether the arbitrators took into consideration EU law. On the other hand, there is the extended review approach, as described by AG Saggio, investigating the correctness of arbitrators’ concrete application of the EU competition law to the merits of the case. The in-depth analysis of this matter will be covered in Chapter 4, here it suffices to mention that although the extended review is less popular than the limited one, national courts do not have yet agreed on a uniform standard of review. To summarise, the two standards of review followed respectively by the US courts and the EU courts are obviously different. The extraterritorial application of antitrust provisions will therefore be affected when international arbitrators have to solve the conflict of law question. This means that in front of a US judge, the application of the EU antitrust law is most likely to be considered satisfactory. While European courts are more strict in pursuing the application of their own mandatory rules by arbitrators and are more likely to pretend the application of EU competition law. Therefore, in order to guarantee the legitimate expectations of the parties when trying to provide an enforceable award, the arbitrator may prefer to apply EU competition law over US antitrust law.138 Such outcome have to be considered also in light of the fact that international arbitration is most likely to face the application of antitrust law to vertical agreements. However, as it has been discussed in the introduction of this Section, the US and the EU antitrust law regimes incur into substantial differences when assessing a violation of competition law through vertical restraints.139 Therefore, it has been argued that the EU competition law ‘could practically become the de facto law governing certain categories of international transactions that are routinely subject to international arbitration.’140 In other words, the EU competition rules on vertical restraints will find global application in arbitration disputes. At first, from the European point of view, the situation described above may 138

Landi and Rogers [2007]. See introduction above. 140 Landi and Rogers [2007]. 139

89

2. ARBITRAL TRIBUNALS’ JURISDICTIONAL ISSUES IN THE APPLICATION OF EU COMPETITION LAW

appear acceptable to protect competition within the EU internal market. However, the limitations of this position appear clearly, when considering that the internal market is part of a truly global economy. The European Union competition law has to face not only the extraterritorial application in transactions with the undertakings from the US, but also with undertakings located in Russia, China, Brazil, India, and others. In a case involving, for instance, the Russian antitrust law, the question is whether a Russian court will easily accept the extraterritorial application of EU antitrust rules. If the answer is negative, then EU undertakings may be deprived from an enforceable arbitral award.

2.2.5

New Challenges to the EU Conflict of Law Approach to Antitrust Arbitration

The previous part of this Section explains why a consistent and objective choice of law rule is necessary to avoid neglecting private parties ex ante expectations about the law applicable to the dispute and also to satisfy the public policy concerns of all the states involved. Moreover, it clarifies that identifying an objective and valid choice of law rule is not enough. This rule needs to be accepted across different jurisdictions and applied uniformly by both private arbitrators and national courts in the phase of the second-look review. Otherwise, as argued by two prominent legal scholars, national judges or arbitrators may be inclined systematically to favour the most severe antitrust regime.141 Given the existing differences among antitrust regimes, this will ax ante affect the outcome of the dispute, as well as the development of international trade and commerce.142 In fact, the presently global economy is deeply influenced by legal, economic and political relations among States. The risk is to incur into retaliation practises by national courts insulating the unwelcome private enforcement of competition law in arbitration and making it ineffective during the phase of award recognition and enforcement in their country. The rapid escalation of a recent case related to Russia seems to fulfil this prediction. 141 142

Landi and Rogers [2007]. Landi and Rogers [2007].

90

2.2.5.1

A Hypothetical Russian Example of Public Policy Exception

In September 2012, the European Commission started formal proceedings to investigate whether Gazprom, the Russian producer and supplier of natural gas, might be hindering competition in Central and Eastern European gas markets, in breach of EU antitrust rules. The European Commission believes Gazprom may have infringed competition in the European markets by restricting the interstate trading of natural gas, limiting the diversification of gas supplies by blocking rival gas-pipeline projects, and pegging the price of natural gas to oil prices in long-term contracts.143 According to the European and the American press, these practises reflect decades-old Gazprom strategy of seeking the advantages of a monopoly, rather than competing under the free-market rules that the EU has been developing, trying to create a single, 27-nation energy market. Allegedly, Gazprom’s business model has long been to buy cheap gas from Central Asian producers that have no alternative route to reach Europe. Gazprom then moves the gas through Russian pipelines and sells it to EU energy companies at twice or three times the purchase price.144 The Commission has concerns that Gazprom may be abusing its dominant market position in upstream gas supply markets in Central and Eastern European Member States, in breach of Article 102 of the Treaty on the Functioning of the European Union. As mentioned above, the Commission is investigating three suspected anticompetitive practises. First, Gazprom may have divided gas markets by hindering the free flow of gas across Member States. Second, Gazprom may have prevented the diversification of supply of gas. Finally, Gazprom may have imposed unfair prices on its customers by linking the price of gas to oil prices. Ultimately, if established, such anticompetitive behaviour would harm EU suppliers and consumers.145 As a matter of fact, the EU gets 36% of its natural gas from Russia. But it is the effective sole supplier to Bulgaria, Estonia, Latvia, Lithuania and Slovakia. According to EU data, it also supplies 82% of 143

See European Commission’s press release on 4 September 2012. Matthew Bryza’s September 18 2012 article published on Bloomberg. 145 See European Commission’s press release on 4 September 2012. 144

91

2. ARBITRAL TRIBUNALS’ JURISDICTIONAL ISSUES IN THE APPLICATION OF EU COMPETITION LAW

Poland’s gas, 83% of Hungary’s and 69% of the Czech Republic’s.146 Nevertheless, in Russia Gazprom’s monopoly in the export of gas to Europe is considered an extremely important matter of public policy. In fact, there was a strong reaction to the Commission’s investigation on Gazprom. Only because the EU Commission opened a formal investigation, given the high value at stake and the geopolitical importance of the matter, Russian leaders have complained in a public outcry. It has been commented in newspapers that the EU is targeting Gazprom because of anti-Russian prejudice. Russian President Vladimir Putin signed a decree which gives Moscow the right to protect its monopoly on natural gas exports from a EU probe. It prohibits strategic companies from disclosing information, disposing of assets or amending contracts without approval from Russian authorities, in case claims are made against them by foreign states or entities.147 No attention was paid to the fact that the conduct of the EU may be considered legitimate and consistent with what has been done not long ago to stop another foreign monopolistic giant from infringing EU competition law, the Washington-based Microsoft Corporation.

2.2.5.2

Overview of the Russian Approach to Antitrust Arbitration

The question is whether such political approach has any legal ground to interfere with the effectiveness of antitrust arbitration. Both, under Article 244(1)(7) of the the Commercial (Arbitrazhnyi) Procedure Code of 2002 (the APC) and under Article 36(1)(2) of the Russian Federation Law on International Commercial Arbitration, a Russian court may refuse to recognise and enforce an arbitral award if it is contrary to Russian public policy (ordre public). If the seat of the arbitration is in Russia, when the court determines that a Russian arbitral award violates Russian public policy, it may also set aside the award.148 Therefore, the means by which Russian courts interpret the public policy clause may be a key factor for parties in international commercial arbitration.149 Public policy, in the context 146

Matthew Bryza’s September 18 2012 article published on Bloomberg. Russian President’s office statement on Tuesday 11 September 2012. 148 Art. 34(2)(2) of the Russian Federal Law on International Commercial Arbitration. 149 Davydenko and Kurzynsky-Singer [2009]. 147

92

of the recognition and enforcement of international arbitral awards, is a broad and vague term which encompasses both substantive and procedural law.150 The High Arbitrazh (Commercial) Court of the Russian Federation (the HCC) formulated the most comprehensive definition of Russian Federation public policy: ‘The international arbitral award can be deemed to violate Russian Federation public policy if its enforcement would result in actions expressly forbidden by law or causing damage to the sovereignty or security of the State, affecting interests of large social groups, being incompatible with the fundamental principles of various States’ economic, political and legal systems, disturbing citizens’ rights and liberties, as well as being contrary to basic principles of civil legislation, such as equality of the participants, inviolability of property and freedom of contract. This definition is not limited to legal aspects. Causing damage to the sovereignty or security of the State or affecting the interests of large social groups is more of a political matter and not necessarily a legal one.’151 The same opinion is followed by the legal literature. Neshatayeva distinguishes public policy stricto sensu and lato sensu. The former limits the public policy notion to fundamental principles of Russian law, whereas the latter also includes morality bases, core religious postulates, and main economic and cultural traditions, whether or not they are expressly established by law.152 However, a review of the case law shows that such broad interpretation is rare in practice. The question now is whether the application of foreign antitrust rules instead of Russian competition law, or the violation of competition law, is a matter of public policy. The Russian legislation does not define how the competition law provisions should be evaluated with respect to public policy. The Russian law on international arbitration and in the APC substantially copied the NYC public policy exception mentioned above. It only refers to a general possibility to deny recognition and enforcement of foreign arbitral awards in case the latter is in 150

In russian, Boris R. Karabelnikov, Problema publichnogo poriadka pri privedenii v ispolnenie resheniy mezhdunarodnognyh kommercheskikh arbitrazhey [Public policy issue on recognition and enforcement of international arbitral awards], 8 Zhurnal Rossiyskogo Prava [Russian Law Journal] 102 (2001). 151 HCC judicial collegium decision of Dec. 6, 2007, No. 13452/07. 152 In Russian, Eshatayeva, N. in Mezhdunarodnoe Chastnoe Pravo i Mezhdunarodnyi Grazhdanskiy Protsess [Private International Law And International Civil Procedure] 560 (2004).

93

2. ARBITRAL TRIBUNALS’ JURISDICTIONAL ISSUES IN THE APPLICATION OF EU COMPETITION LAW

violation of the public policy provision of the Russian Federation. Also in the Russian legal literature there is no direct answer to the question whether it is necessary to associate the violation of competition law to a violation of Russian public policy matters. Focusing on the case law, on 4th of October 2006 the Federal Arbitrazh Court of the Moscow Circuit adopted a resolution on the case n. A40-46077/05-25-2006 on the dispute between OAO Tambovskaya energosbytovaya kompaniya (TEK) and the OAO Tambovkurort. The two parties signed a contract for the supply of energy which includes an arbitration clause referring any future dispute to the dispute resolution body of the OAO Unified Energy System of Russia. After a competition law dispute between the parties has been solved by this arbitration court, the Tambovkurort appealed to the first instance Arbitrazh Court of the Moscow, requesting the annulment of the award. The case eventually ended up to the Federal Arbitrazh Court of the Moscow Circuit. One of the arguments used to request the annulment was the following: Tambovkurort claimed that the arbitral tribunals solved the dispute in violation of the Russian competition law.153 It also claimed that the decision violates the fundamental principles of the Russian law including the one about limitation of monopolistic activities. The Appeal court noticed that, even the wrongful application by the arbitral court of substantive law, including the competition law, in itself (per se) cannot justify the annulment of the arbitral award on the base of a violation of fundamental principles of law (public policy). In fact complaints upon the merits of the dispute have to be made in front of the arbitral tribunal and not in front of a national court. Hence the appeal court focused its attention on the point that competition law should be applied within an arbitral proceeding. In practise, the appeal court indicated that the application of competition law as substantive law provision should be decided by arbitration courts on the merits and do not need to be reviewed by a national court. The appeal court states that even the wrongful application per se does not mean the violation of fundamental principles of Russian law, namely public policy. Therefore, in this case, the appeal court did 153

Law of the RSFSR on Competition and Limitation of Antitrust Activities on Commodity Markets, dated 22 March 1991, Articles 27 and 28; Rules on the Resolution of Cases on the Violation of Antitrust Legislation, adopted by Order No. 53 of State Committee of che Russian Federation on State Antitrust Policy and Support to New Economic Structures, dated 12 May 1994.

94

not link the violation of competition law to the violation of Russian public policy. As it has been argued by Gurkov, the court followed a more liberal orientation with respect to the application of competition law in arbitration proceedings, in comparison to the CJEU’s Eco Swiss decision.154 However it should be noted that the appeal court did not deal with a case of recognition or enforcement of a foreign arbitral award applying foreign competition law but with a domestic arbitration applying Russian competition law. Nevertheless this case may lead us to believe that also the recognition and enforcement in Russia of foreign antitrust arbitral awards will not be extremely problematic. Such impression will rapidly change analysing the underlying implications of a case-law of the Supreme Arbitrazh (Commercial) Court on a choice of law matter.155 Although the Delta Vilmar v. Efirnoe case does not regard directly competition law, as argued by Nigmatullina, the Russian court substantially stated that an arbitral tribunal’s conflict of law determination is subject, as a matter of procedure, to national courts’ control.156 Consequently, an incorrect determination of substantive law can amount to a violation of public policy in Russia. On the contrary, courts in the majority of Western jurisdictions disregard arguments that arbitrators’ incorrect determination of the conflict of law issues violates the agreement of the parties.157 Moreover, usually such issues are regarded as substantive ones and are subject to very limited judicial review, if any. On the contrary, in the case at hand, the Supreme Arbitrazh (Commercial) Court vacated the decisions of the lower court that had set aside the arbitral award, instead of choosing the straightforward solution. That is to declare, as most developed jurisdictions do, that the determination of the applicable law constituted the substance of the parties’ dispute and thereby could be subject to no judicial review. The Supreme Arbitrazh (Commercial) Court chooses a different path. It examined the merits of whether the choice of law issues had been correctly determined by the arbitrators. Therefore, the lower courts’ decisions were 154

Gurkov [2013]. Russian Supreme Commercial Court case A27-4626/2009 OJSC Efirnoe (hereinafter Efirnoe) and a Ukrainian seller, LLC Delta Vilmar CIS (hereinafter Delta Vilmar). 156 for International Arbitration [2012], p. 164. 157 for International Arbitration [2012], p. 164. 155

95

2. ARBITRAL TRIBUNALS’ JURISDICTIONAL ISSUES IN THE APPLICATION OF EU COMPETITION LAW

annulled because the application by the arbitral tribunal of the CISG as part of the Russian law did not amount to a clear violation of public policy. Thus, even though the Supreme Arbitrazh (Commercial) Court came to the correct conclusion, it did so for the wrong reasons and following the wrong argumentative paths. It opened what can be called a “pandora box”, which is the possibility to review whether the arbitral tribunal have chosen to apply the correct law. In this way, the approach taken would justify, in future cases, the same inquiry by all the three degrees of Russian courts. Unfortunately, this demonstrates that allegations of violation of public policy on the ground of incorrect determination of the substantive law (including competition law) still can succeed in Russia. The result of such review activity would jeopardise the enforceability of international arbitral awards and consequently impair the effectiveness and the functionality of international arbitration as a tool for the private enforcement of competition law. Both legal certainty and private parties expectations on dispute resolutions in the context of international business transactions could be adversely affected by this attitude. To conclude, this Chapter deals with three major unresolved jurisdictional issues of antitrust arbitration. These issues can be summarised as follows: (i) the suitability of arbitral tribunals to solve antitrust disputes; (ii) the requirements that a contractual arbitration clause needs to satisfy to validly devolve to arbitrators the power to solve an antitrust dispute; and (iii) what choice of law rule should guide the arbitrators’ decision on the law applicable both to procedural issues and to the merits of a case. Along with these topics the Chapter provides a general legal analysis of antitrust arbitration, defining the legal framework within which this research project operates. This legal framework extrapolates and highlights the key elements that must be taken into consideration when evaluating the efficiency of alternative legal solutions provided by lawmakers in this subject matter.

96

Chapter 3 The Jurisdictional Issues of Antitrust Arbitration Under the Lens of Game Theory The previous Chapter provides a legal framework describing three major unresolved jurisdictional issues in antitrust arbitration. That legal analysis aims to outline the open legal questions as well as the existing procedural law rules that come at stake in each subject matter. Moreover, the emerging legal framework allows us to extrapolate the key elements that must be taken into consideration when evaluating the efficiency of alternative legal solutions provided by lawmakers. This Chapter employs game theory as a descriptive tool that can formally clarify the failures of the decision-making process discussed in the legal analysis above.1 Only the previously defined legal framework could have provided the background that is necessary to justify the formal description of strategic situa1

The reader may wonder whether game theory should be used to represent a scenario that could be told in words or by means of case studies. Game theory can be a powerful descriptive tool that provides specific advantages over a pure legal analysis. A game theory framework facilitates a deeper understanding of the players’ incentives, their interests and the interrelated effects of their actions on each other. In a game theory model, all these elements are structured and organised into a clear-cut manner. Therefore, such an analysis facilitates the comparison between possible alternative legal rules that can be used to solve failures in the legal system.

97

3. THE JURISDICTIONAL ISSUES OF ANTITRUST ARBITRATION UNDER THE LENS OF GAME THEORY

tions that is performed in this Chapter. Only in light of that legal analysis it is possible to understand why the matrices and the game trees of this Chapter have exactly the form and the payoffs described below. The legal analysis above, with its descriptions of case-law, legal provisions and practical examples, guarantees that the game theory assumptions made in this Chapter are connected to real world scenarios. Therefore, this Chapter will constantly refer to relevant parts of the legal analysis of the previous Chapter to explain and support the arguments used herein. Regarding the structure of this Chapter, it is divided in 5 Sections. The first one, called ‘correcting a less desirable outcome through competition law private enforcement’, is a building block. It is a premise that helps to clarify the strategic problem that private enforcement of competition law aims to address. Moreover, as we said, it is a useful building block. This first Section starts briefly overviewing the results of the Law and Economics literature on vertical restraints and abuse of dominant position and functions as a guiding light during the progression of this economic analysis. Our goal, here, is to put into a game theory perspective the origin of an anticompetitive relationship between two business parties. In other words, there is a bargaining problem that can lead two economic actors towards a less desirable outcome. Private actions for competition law damages are a tool that can function as a cure aiming to correct this problem. However, private antitrust enforcement was not conceived taking into consideration the features of international commerce. For international business parties a possible alternative to private enforcement is to solve contractual antitrust disputes resorting to an impartial international arbitrator. The second Section, ‘An argument against limitations to antitrust arbitrability’, elaborates on the reasons that, at first, lead the lawmaker to intervene limiting the private parties’ autonomy to solve contractual disputes for antitrust damages in front of international arbitrators. Elaborating on the shortcomings of this approach, this second Section supports a more extensive interpretation of the arbitrability concept, resembling the one recently adopted in France and the US.2 2

See previous Sections above.

98

Incidentally, the third Section tries to provide a useful argument that answers the question: ‘Should competition law opt-in or opt-out of an arbitration clause?’. Recently several legal systems have adopted different solutions on the question whether competition law should be explicitly mentioned in an arbitration clause to consider that the parties wanted to devolve to arbitrators the solution of their antitrust disputes. It is a key question because it defines the scope of arbitration clauses and dictates a rule on how should they be drafted.

Although antitrust arbitration has the potential to overcome the weaknesses of national private enforcement, in the context of international commerce it can create further challenges putting at risk the very functioning of private enforcement of competition law. Within the forth Section, ‘Not all arbitration clauses are alike’, the concerns expressed by those who argue against antitrust arbitration are not left unheard. Contractual arbitration clauses are the result of private bargaining. Hence, they can be affected by the same incentives leading economic players to reach a less desirable outcome for competition law enforcement. This Paragraph describes what happens into a conflict of law scenario when the subjective arbitrability and the validity of an arbitral award are tested under different choice of law rules

In the last Section, we try to provide ‘A possible solution to the conflict of law issues in antitrust arbitration’. In light of the absence of a unique and consistent choice of law rule to guide international arbitration and national judges towards the solution of the conflict of law problems in antitrust arbitration, we propose an alternative solution capable to satisfy the needs of legal certainty in the context of international commerce.

99

3. THE JURISDICTIONAL ISSUES OF ANTITRUST ARBITRATION UNDER THE LENS OF GAME THEORY

3.1

Correcting a Less Desirable Outcome Through Competition Law Private Enforcement

In order to analyse the legal problems mentioned above using the game theory lens, this Section reduces the legal problem underneath the antitrust private enforcement of vertical restraints to its essentials. In light of the methodology described in the preamble of this work, and of the notions presented in the introductory Chapter, one can think about the private enforcement of competition law as a game between two players a manufacturer and a distributor that have signed a vertical agreement. However, before delving into the game theory analysis it is worth mentioning the Law and Economics literature that discusses the antitrust problems underneath vertical agreements. Manufacturers need to distribute their products to customers and to provide necessary services or complementary products capable to maximise their profits. There are two possible alternative ways to achieve this result. A manufacturer can (a) directly sell his products to the final customers. Alternatively he can (b) use independent distributors and retailers. The former option (a) implies the vertical integration of a firm. The integrated firm’s management would decide how to optimally organise every element of the production-distribution chain to maximise the total profit. The latter option (b) relies upon market transactions between two firms, one operating at the upstream level and the other at the downstream level of a production chain. Since the two firms are not integrated, the upstream firm sells products to the downstream firm. This means that a manufacturer does not compete against his retailer. A retailer can potentially provide access to the downstream market as well as complementary products or services that can increase the manufacturer’s profits. However, the two firms need to coordinate their actions to maximise their joint profits. Otherwise, without the appropriate coordination, one of the two firms can undertake selfish decisions maximising its own profit to the detriment of the counterparty’s profit. In other words, each firm’s decision affecting price, quality, retail services, and so forth can generate the so-called spill-over effects on their joint profits. To avoid this undesirable outcome the two entities govern their relationship through a contract

100

called vertical agreement. Vertical agreements are usually long-term contracts that decide and regulate many conditions of a vertical relationship. They can also incorporate contractual clauses that put various restrictions to the firms’ activities, i.e. vertical restraints. It is straightforward to think about vertical restraints as contractual clauses imposed by the upstream manufacturer on the downstream retailer. However, this is not always the case, sometimes the opposite can be true. A retailer may have the stronger bargaining power and impose vertical restraints on its upstream supplier.3 However in this brief overview we will focus on the common case of an upstream firm imposing restrictions on the activities of its downstream counterparty. Vertical restraints can be designed in different forms. They can aim to control the retail price (so-called Resale Price Maintenance or RPM), the quantity sold (minimum quantities), the territorial exclusivity of each retailer, the purchases of additional products (so-called tying), and so forth. In some circumstances vertical restraints can enhance competition and welfare (internalising externalities and providing the retailers with welfare maximising incentives) while in others they can have the opposite effect (enhancing market power and reducing competition). To make things more complicated, in many cases the effect of vertical restraints does not directly depend on their form. In order to clarify this point we will overview, below, several instances of vertical restraints that can create pro-competitive welfare enhancing incentives. Retailers have to compete in the downstream market against other retailers,4 hence they have the incentive to reduce the products prices or the quality of connected services (allowing some retailers to free-ride on the services provided by other retailers generating horizontal externalities).5 This can reduce the manufacturer’s profit but also push other retailers out of the market reducing the competition in the downstream market.6 Manufacturers can prevent such a re3

Foer [2004];Grimes [2004]; Noll [2005]. As well as, symmetrically manufacturers compete in the upstream market against each other. 5 Telser [1960]. 6 In other words since retailers cannot fully internalise the profits from an investment they 4

101

3. THE JURISDICTIONAL ISSUES OF ANTITRUST ARBITRATION UNDER THE LENS OF GAME THEORY

duction of profits and competition including restraints into the vertical agreement that they sign with retailers.7 Another example of the same incentives leading towards a lower quality final product can be imagined in a franchise contract. If a franchisee can use less expensive inputs, he would increase his profits selling lower quality goods under the franchisor’s brand. This conduct can be easily corrected with properly designed vertical restraints.8 A further example of spill-over effects that can be avoided using vertical restraints involves a manufacturer that does not want to invest into retailers’ shops (for instance training the dealers’ personnel or improving the quality of their facilities). Vertical restraints can protect his investment from the retailers free riding and they can also guarantee that his investment does not benefit also the competitors’ products sold in the same shops. The simple solution that can protect the manufacturer’s investment is an exclusivity clause in the vertical agreement with the retailer. An additional problem that vertical restraints aim to solve is the so-called double marginalisation problem. This is a typical problem that occurs when both the upstream firm and the downstream one have market power. Without vertical restraints the manufacturer and the distributor cannot coordinate appropriately and will set the product price too high.9 Thus, with higher prices the quantity sold will not reflect the amount capable to maximise their aggregate profits. The theoretical explanation of this outcome can be dated back to the work of Carnout in 1838.10 The solution to the double marginalisation problem can be provided setting a maximum resale price, a minimum quantity requirement or a two-part tariff. It is worth noting that all of the above described vertical restraints can be defined as efficiency enhancing because they decrease the prices that customers pay while increasing the profits for the firms. will decrease the amount of such investment. See Mathewson and Winter [1984]; Winter [1993]. 7 Klein and Murphy [1988]. They propose to use minimum resale prices and exclusive territories to solve this problem. 8 In this example a tying clause could prove to be appropriate, forcing the franchisee to purchase all the inputs from the franchisor or from a previously approved supplier. 9 Holmstrom [1982]. 10 A successive complete analysis comes from Spengler [1950].

102

These few examples clarify why, in some circumstances, vertical restraints can have pro-competitive effects. However in other circumstances the opposite can be true. For instance, when one of the involved parties has a sufficiently strong market power, vertical restraints are more likely to produce anticompetitive effects.11 In other words, when in the supply chain there is a company that has enough bargaining power it has the incentive to introduce vertical restraints capable to adversely affect its direct competitors. There are two classic theories about vertical restraints that can generate anticompetitive effects.12 The first theory (the so-called input foreclosure theory) is not common because it requires that a downstream firm has sufficient market power to force its upstream counterparty to increase the costs of the products that the latter sells to the downstream competitors.13 Examples of vertical restraints imposed by distributors include: exclusive supply obligations, reciprocal buying, refusal to buy, enforced sale-or-return, delayed payments, and so forth.14 Under the second theory, a manufacturer could use vertical restraints to induce retailers to purchase fewer goods from the manufacturer’s competitors (the socalled customer foreclosure theory). These practices can be perpetrated through specifically designed vertical restraints that can assume many forms, for example, territorial exclusivity clauses, a specific rebates structure, quantity forcing, price fixing, refusal to supply, tying (including bundling and full-line forcing), selective distribution restrictions (defining the group of buyers to whom goods can be sold). For instance we could think about the case of a manufacturer that has introduced an exclusivity clause with its whole network of retailers. If this network involves 11

In fact, the so-called Block Exemption Regulation considers vertical agreements as unharmful when the involved parties’ market share is less than 30%. See Commission Regulation (EU) No 330/2010 of 20 April 2010 on the application of Article 101(3) of the Treaty on the Functioning of the European Union to categories of vertical agreements and concerted practices, OJ L 102 of 23.4.2010. 12 Cooper et al. [2005]; Rey and Tirole [2007]; Hart et al. [1990]; Salinger [1988]; Reiffen and Kleit [1990]. 13 Foer [2004]; Grimes [2004]; Noll [2005]. 14 The list is non-exhaustive and the meaning of these clauses has not been explained because an analysis of vertical restraints is out of the scope of this work. However in the cited literature describes and analyses these clauses.

103

3. THE JURISDICTIONAL ISSUES OF ANTITRUST ARBITRATION UNDER THE LENS OF GAME THEORY

almost all the retailers, it could be excessively costly for the upstream competitors to reach the final customers. In a similar context, vertical restraints could reduce competition, create a monopoly or create a sufficiently high entry barrier that protects a monopolist from the competition of newcomers.15 Furthermore there is also another possible anticompetitive effect generated by vertical restraints. They could be used to simplify the creation and continuation of cartels.16 For instance, an upstream firm that includes a minimum price clause in all its vertical agreements would facilitate the creation of a cartel between retailers that can enforce the application of monopoly prices.17 Moreover vertical restraints could be also used by manufactures to establish and govern cartels. In fact, upstream firms can use a common distributor as a tool to coordinate on prices or quantities produced.18 For instance, exclusive distribution agreements could be employed to facilitate artificial territorial segmentations as well as the stability of a cartel. Alternatively, the simple use of resale price maintenance clauses could favour the detection of deviations from a previously agreed collusive practice.19 To summarise this overview, vertical restraints serve as a contractual substitute for vertical integration. They can have either positive effects or negative effects on competition and welfare. As an author puts it: ‘the appropriate treatment of vertical restraints may be the most controversial subject in antitrust’.20 Another author clarifies that: ‘no simple conclusion can be drawn whereby any particular type of vertical restraint (territorial, restrictions, tie-ins, RPM, etc.) will inevitably improve economic efficiency or reduce it’.21 The economic and competition effects can change in different contexts.22 In other words, notwith15

This is a bottom-line simplification of a sophisticated legal and economic theory that has specific requirements. See Whinston [1989]; Carlton and Waldman [1998]; Nalebuff [2004]; Reiffen and Kleit [1990]. 16 See Jullien and Rey [2007]; Ornstein [1985]. 17 See Telser [1960]. Critics to this argument state that it should be explained why manufacturers would have the incentive to support such a cartel. However, it has been clarified that the manufacturers could use a two-part tariffs clause to gain part of the downstream applied higher cartel price. 18 Bernheim and Whinston [1985]; O’Brien and Shaffer [1997]. 19 Jullien and Rey [2007]; Hastings and Gilbert [2005]. 20 O’Brien [2008], p. 40. 21 Rey [2008]. 22 Rey [2008]: ‘if the market structure (level of concentration, conditions of entry, market

104

standing the fact that vertical restrains are widespread, when one of firms has sufficient market power, assessing whether they are pro-competitive or anticompetitive can be a controversial matter. To quote Margaret E. Slade, ‘many economists think that their [vertical restraints] principal role is to enhance efficiency and should therefore be viewed as beneficial, whereas others believe that their primary purpose is to increase market power and should therefore be considered pernicious. Moreover, conflicting academic attitudes are mirrored in changing and inconsistent antitrust policies towards vertical restraints.’ The author also adds that ‘these attitudes and classifications often differ by restraint within a jurisdiction and time period, by jurisdiction within a time period, and over time within a jurisdiction.’23 Therefore, even if there is an established literature on vertical restraints, on exclusive contracts and on the abuse of dominant position it may be complicated for a business party, especially a small retailer, to assess whether a vertical agreement offered by a manufacturer has pro-competitive or anticompetitive effects.24 In light of this digression on vertical restraints it is now clear that in a vertical relationship manufacturers and retailers have two very different positions. In order to continue with the description of this relationship in a game theory setting, we have to take into consideration the two parties’ contrasting incentives. For the sake of simplicity we start the game theory analysis with a simple two players game. Player I is a manufacturer that has a dominant position in the upstream dynamics, and so forth) ensures vigorous competition among rival vertical structures, vertical restraints are unlikely to harm economic efficiency or reduce competition. Conversely, in less competitive markets the risk is much greater that vertical restraints can be used to reduce competition or otherwise reduce economic efficiency. 23 Slade [2008]. 24 It is worth reminding the reader that, as clarified in the Preamble Section of this work, it is out of the scope of this research project to study the competition law substantive rules. This work focuses on procedural Law and Economics issues that come at stake during the private enforcement of competition law in arbitration. The Industrial Organisation literature mentioned above deals with competition policy issues that are addressed by substantive law rules. On the contrary, enforcement policy issues are addressed by procedural law rules. The private enforcement of competition law either in front of a national court of in front of an arbitral tribunal is simply an enforcement tool. We do not believe that procedural law rules should be used to address issues that do not have a procedural nature. Therefore, we believe that in this work there is no need of a in-depth account of the Law and Economics literature on vertical restraints. In this research project antitrust substantive law is taken for granted.

105

3. THE JURISDICTIONAL ISSUES OF ANTITRUST ARBITRATION UNDER THE LENS OF GAME THEORY

market, or that simply has a strong bargaining power.25 He can be identified with Gazprom or Coca-Cola in our previous examples.26 This means that Player II is a retailer capable to distribute the manufacturer’s products in the downstream market. Our game theory analysis starts with the first interaction between these two players: the signing of a vertical agreement. This vertical agreement can be profitable for both of them but they have a different preferences structure on how it should be designed. The manufacturer (Player I ) has the incentive to use the retailer (Player II ) to reinforce and consolidate his market power, in other words to pursue an anticompetitive agenda.27 Meanwhile the retailer has the incentive to sign a vertical agreement that generates profits for her, although it may be difficult for her to determine ex ante whether the vertical agreement offered by the manufacturer is pro-competitive or anticompetitive.28 Obviously in the latter case, if the manufacturer succeeds in reducing the upstream competition (for instance enforcing a customer foreclosing vertical restraint), if the manufacturer becomes a monopolist or a dominant firm the retailer’s mark-up may be severely reduced. Given this description of the players’ preferences, each player will act in light of the information available to him/her. As explained at the beginning of this Section, Player I is in a better position to evaluate whether the vertical restraints that he is offering to Player II can lead to an anticompetitive or pro-competitive outcome. On the contrary, Player II cannot fully evaluate whether in that spe25

It is possible to generalise the model and include also the case when a retailer has a dominant position in the downstream market. In this case the two parties will invert their respective positions. However this Chapter focuses on the more frequent case of a manufacturer with bargaining power because this type of scenario is at the centre of the case-law described in the legal analysis above. 26 For the sake of clarity, throughout the game theory descriptions of this Chapter we use the male pronoun ‘he’ when referring to Player I and the female pronoun ‘she’ when referring to Player II. 27 We assume that Player I has already taken into consideration the public enforcement of competition law. This means that his cost-benefit analysis on breaching competition law provided is positive result. At this point of our analysis private enforcement has not yet been introduced. 28 Moreover it may even not be economically justified for the retailer to invest ex ante into a complex antitrust analysis to solve this ambiguity. In case ex post it becomes clear that the contract is anticompetitive it is in the social welfare advantage if the damaged party has the right to nullity the effects of this contract.

106

cific context the offered vertical agreement is pro-competitive or anticompetitive. She only knows her local market share, her business strategy, break-even point and the margins of profit that she can expect to realise given the market power of Player I. In light of this information set, the game can be defined as a sequential game. The first player to act is Player I who offers a vertical agreement including either pro-competitive vertical restraints or a anticompetitive ones, AP I = P ro; Anti. According to our initial assumption Player II does not have the necessary information to determine ex ante whether the vertical agreement offered by Player I is pro-competitive or anticompetitive. After she has been offered the contract, the two actions that Player II can undertake are the most obvious, she can sign the contract or not sign the contract, AP II = Buy; N otBuy. Before moving to the payoffs determined by each action for each player, the figure below provides a representation, in extensive form, of the game. Player I P ro

Anti

Player II Buy

{..;..}

Player II

N otBuy

{..;..}

Buy

{..;..}

N otBuy

{..;..}

In the sequential game above, each player makes one decision and they undertake two actions. Player I is the first to act; when Player I acts he does not know how Player II will act. However the dashed line means that when Player II acts he does not know how Player I has acted in the previous stage of the game; he does not know the exact node that has been reached. Therefore, although in reality the actions are sequential, the players’ specific incomplete information

107

3. THE JURISDICTIONAL ISSUES OF ANTITRUST ARBITRATION UNDER THE LENS OF GAME THEORY

sets allow us to study this game as if it was a simultaneous game and represent it using the strategic form or normal form.29 This game description is not complete without the players’ payoffs. Since we are not interested in a quantitative analysis but only a qualitative one, we decided to assign specific numerical payoffs to each strategy. Each player’s payoff reflects his preference ranking of that specific strategy. Keeping in mind the players’ preferences description above, we can order all the possible strategies of this game. For Player I the best case scenario is that an anticompetitive vertical agreement is signed by Player II, so his payoff for (Anti; Buy) can be ranked to the highest position assigning to it the value 3. Player I will profit less if a procompetitive contract is signed, so his payoff for (P ro; Buy) can be ranked lower at 2. When Player II does not sign a pro-competitive contract Player I does not realise any profit but he is better-off in relation to having an anticompetitive contract refused by Player II.30 Therefore Player I ’s payoff for (P ro; N otBuy) is 1 and his payoff for (Anti; N otBuy) is equal to 0. To summarise, the payoffs that we have assigned to Player I reflect the order of his preferences, so: (Anti; Buy) > (P ro; Buy) > (P ro; N otBuy) > (Anti; N otBuy). Respectively for Player II the best case scenario is to sign a pro-competitive vertical agreement, hence the payoff for (P ro; Buy), is the highest value 3. Since, as described above, in the long run Player II will be damaged by an anticompetitive contract the payoff for (Anti; Buy) ranks lower at the value 2. Conversely not signing an anticompetitive contract does not generate any profit but it is a better case scenario than refusing to sign a perfectly legitimate pro-competitive vertical agreement that could have generated a profitable long-term business relationship. Hence, we can assign to Player II ’s payoff for (Anti, N otBuy) the rank value 1, while the payoff for (P ro, N otBuy) will receive the lowest value 0. To summarise, the payoffs that we assigned to Player II reflect the order of his preferences, so: (P ro; Buy) > (Anti; Buy) > (Anti; N otBuy) > (P ro; N otBuy). 29

See Rasmusen [2006], pp. 41-42. This statement is justified in light of the lost opportunity of a higher potential profit under an anticompetitive contract in comparison with the lower profit under a pro-competitive one. 30

108

The described game is represented below in a strategic form. Payoff Matrix Player II

Player I

Buy

Not Buy 3

Pro 2

0 1

2 Anti

3

1 0

Even from an elementary game theory analysis of the players strategic interaction in this game, we can say that there is a unique outcome. Namely, an anticompetitive agreement will be signed by the parties. Player II has a dominant strategy. Independently of what Player I does, Player II is better-off if she take the action (Buy), signing the contract. Knowing this, the best strategy for Plater I is to offer an anticompetitive contract. Therefore the static equilibrium of this game is the strategy (Anti; Buy). In order to bring this game theory description closer to reality we have to consider that such a game can be played multiple times. However we do not believe that a dynamic game theory analysis would be appropriate. In fact usually vertical agreements are long-term contracts and Player I ’s reputation will be affected by the first game played between the parties.31 After it becomes clear that the vertical restraints offered by Player I have anticompetitive effects, Player II will expect that Player I ’s offer to renew the vertical agreement will also be anticompetitive. Moreover if the anticompetitive vertical restraints have indeed succeeded in reducing the level of competition in the upstream market, it is likely that Player II would not have many options. Either she signs renewing the vertical agreement or she may not have any viable alternative option to sell that product in the downstream distribution market. For both social welfare and Player II, the outcome of this game is not the 31

See the literature on Industrial Organisation above.

109

3. THE JURISDICTIONAL ISSUES OF ANTITRUST ARBITRATION UNDER THE LENS OF GAME THEORY

desirable one, because it leads to the stipulation of anticompetitive contracts. It is not within the scope of this research to discuss whether and why a different outcome is preferable, nor whether and why Player I should not be allowed to profit in this way from his dominant position in the market. What matters, for this research, is the role of the private enforcement of competition law. It is worth noting that, as clarified in the Preface Section of this work, a comparison between private and public enforcement is out of the scope of this work. Nevertheless the reader may wonder if those circumstances producing anticompetitive contracts would trigger the attention of competition authorities. Competition authorities would indeed attempt to stop any such infringement of antitrust law, if they were able to monitor all existing vertical restraints signed by firms with market power. However such enforcement system is unlikely to be effective and it will most certainly be too costly for competition authorities. Moreover given that public enforcement does not provide compensation for the damaged party, the latter may not always have the sufficient incentive to intervene signalling potential antitrust violations to the public authorities. Conversely the private enforcement of competition law has the potential to provide the desirable solution to this problem. It can provide retailers with the incentive to start private actions and signal to competition authorities a breach of antitrust law. Precisely on the basis of this rationale legal systems have introduced the private enforcement of competition law, as discussed in the Preface Section of this research project.

In fact within the framework of procedural rules defining a private enforcement system, the damaged party of an anticompetitive vertical agreement has the right to start private actions for damages (the so-called use of competition law “as a sword”) or to set aside an anticompetitive clause (the so-called use of competition law “as a shield”). Therefore, when Player II can enforce her competition law rights, the payoff matrix of the game above changes. Under an effective private enforcement system, the parties can autonomously reach an alternative equilibrium than the one of the game above. This new game produces ex ante a more desirable outcome without the intervention of a public authority. Under an effective private enforcement system Player I is not supposed to have the incentive to offer anticompetitive contracts. When an adjudicator, either na-

110

tional court or arbitral tribunal, decides that Player I has breached antitrust law and has to pay damages and litigation costs, Player I ’s order of preferences in this game changes. An effective private enforcement system should modify the payoffs for (Anti, Buy). The granted amount of damages and enforcement costs should downgrade for Player I the strategy (Anti, Buy) from being the best action to being the second best action with payoff 2 ranked lower than the payoff for (P ro, Buy) which will be then equal to 3. Payoff Matrix with Private Enforcement Player II

Player I

Buy

Not Buy 3

Pro 3

0 1

2 Anti

2

1 0

Therefore only the left side of the payoff matrix changes for Player I, crating a new equilibrium based on the dominant strategy for both players (P ro, Buy). As described above, this preferable outcome can be achieved allowing Player II : (i) to start follow-on actions for damages after Player I has been fined by the European Commission, (ii) to start direct actions for damages, or (iii) to trigger the competition law defence in case Player II breaches the anticompetitive contract.

The result above is not as obvious and expected as one could imagine. This line of reasoning has not always been considered appropriate. The argument used by the opponents of antitrust private enforcement is based on the claim that it is “unjust” and counter-intuitive that the party who has signed and agreed upon an anticompetitive contract should be paid damages on the ground of a competition law violation perpetrated through that agreement. In our example, this legal argument would make Player II co-responsible for the violation of the EU

111

3. THE JURISDICTIONAL ISSUES OF ANTITRUST ARBITRATION UNDER THE LENS OF GAME THEORY

competition law because she signed the anticompetitive contract. Under this line of reasoning, it could be even argued that Player II knew of should have known the anticompetitive nature of the agreement. Therefore she may even risks to be fined by the antitrust authority. Obviously this approach is not the one followed by the EU legal system. Using our game theory description, we briefly elaborate below on this alternative legal rule showing the shortcomings of a similar approach. Under this rule, if Player II signs an anticompetitive contract she cannot receive any compensation for damages and moreover she is facing the additional risk of fines. Either bringing a private claim for damages or not, she may have to pay a fine or even only the costs of defending herself in front of the antitrust authority. Therefore assuming that Player II is risk-adverse and assuming that the expected fine will be higher than the expected profit generated by the anticompetitive agreement, Player II ’s order of preferences changes. Her payoff for (Anti, Buy) should be the lowest value 0. Not signing a pro-competitive contract will not generate profits but will certainly not cause losses either. Hence her payoff for (P ro, Buy) can be assigned equal to 1. Player II would be even better-off if she does not sign an anticompetitive vertical agreement because she avoids a potentially detrimental scenario. This means that the payoff for (Anti, N otBuy) would be ranked lower than (P ro, Buy) but higher than (P ro, N otBuy), so equal to 2. Conversely Player I’s position does not change from the payoff matrix without private enforcement (represented in the first game description above). Under this legal rule he does not have to pay any damages and the risk of being fined by the antitrust authority was already incorporated into his initial order of preferences.32 What follows is the new strategic-form game. The payoff matrix corresponds to the legal rule according to which Player II should not receive compensation for damages, but should be co-responsible of a competition law violation in the eyes of the antitrust authority: 32

An effective public enforcement system should set the expected fine of a competition law breach higher than the expected profits from anticompetitive vertical restraints. However if this was the case, this game would have never been played because a rational Player I would not consider the action (Anti) available.

112

Payoff Matrix with Fines for Player II Player II

Player I

Buy

Not Buy 3

Pro 2

1 1

0 Anti

3

2 0

Such a game is defined in game theory as a discoordination game.33 In this scenario there is no unique equilibrium without repeated games.34 Although, as mentioned above, in long-term contracts there is not much room for mixed strategies, it is worth mentioning what we believe would have happened if it was a viable option. Including repeated games and mixed-strategies the parties may reach an equilibrium. However fewer contracts will be signed in such a scenario in comparison to the contracts signed when private enforcement of competition law is available. When fewer vertical agreements are signed, fewer business transactions occur and this will adversely affect the social welfare. Conversely, the private enforcement of competition law will force the parties to play the different game described above that reaches a desirable outcome. In summary, private antitrust enforcement can provide business parties with the incentives to make them shift from a scenario where they sign anticompetitive vertical agreements, toward a scenario where they sign pro-competitive vertical agreements. In other words an effective antitrust private enforcement system can lead the parties toward a more socially desirable equilibrium (they have an additional incentive to comply with competition law). A private antitrust enforcement system is effective only if the parties have no viable way to prevent 33

Rasmusen [2006] p. 79. Rasmusen [2006] p. 79; the contribution of John Forbes Nash, Jr. in his 1951 article ‘NonCooperative Games’ was to define a mixed strategy Nash Equilibrium for any game with a finite set of actions and prove that at least one (mixed strategy) Nash Equilibrium must exist in such a game. 34

113

3. THE JURISDICTIONAL ISSUES OF ANTITRUST ARBITRATION UNDER THE LENS OF GAME THEORY

the application of competition law rules in private actions for damages. The next Section of this Chapter analyse the procedural rules that can make effective the private antitrust enforcement in arbitration.

3.2

An Argument Against Limitations to Antitrust Arbitrability

The private antitrust enforcement can achieve the results described in the previous Paragraph, proportionally to the effectiveness and the credibility of the compensation threat directed to Player I. If an anticompetitive contract is signed by the parties at the moment T0 , Player II will discover whether it is anticompetitive or not only in a different moment of time, T1 (e.g. after the decision of the European Commission to fine Player I, or after the contract or part of it is performed by the parties). Therefore, at T1 , there should be no limitation or impairment in Player II ’s ability to initiate a private action for damages.35 As argued above in the introductory Chapter, deciding to solve a dispute in arbitration is a legitimate expression of the parties’ autonomy. However the legal analysis of the previous Chapter clarifies that on the basis of his stronger bargaining power, Player I could exploit arbitration to circumvent the application of antitrust rules. Lawmakers believed that business parties may draft arbitration clauses with the intent to take away antitrust disputes from the jurisdiction of national courts. Aiming to prevent this conduct, at first, lawmakers denied arbitrability of competition law disputes, or in general the arbitrability of certain kinds of contracts (i.e. distribution agreements in Belgium, in Saudi Arabia, and in the Emirates). Many reasons can be used to justify limiting the arbitrability of certain disputes, but a game theory analysis of this problem can show that in the field of private antitrust enforcement this enforcement policy may not provide a satisfactory result. The game that is at the cornerstone of this Section has the same players 35

This is exactly the rationale underneath the subjective arbitrability discussed in the legal analysis above.

114

of the previous games, Player I and Player II. Unlike the previous games, this time they play under perfect information. Since in the previous Paragraph we concluded that in any case Player II ’s dominant strategy is to sign the offered vertical agreement, this game assumes that the contract was signed.36 However, in this game there are two different moments of time which determine the order of play; also the players’ action sets are different. In T0 , when Player I is drafting the vertical agreement, he has two possible actions AP I = C 0 ; C − . The action C 0 means that he abstains from including into the contract any arbitration clause and the dispute will be decided by a national court through ordinary litigation.37 The action C − means that Player I includes into the vertical agreement an arbitration clause that limits the effectiveness of antitrust arbitration, making it impossible or extremely difficult for Player II to enforce his antitrust rights. In the previous Chapter we referred to this case when discussing anticompetitive arbitration clauses.38 In T1 , when Player II discovers that she has certain competition law based rights, she will have to choose between a set of two possible actions: AP II = S; B. The action S means that she sues Player I for damages (the so-called use of competition law “as a sword”). The action B means that she breaches the contract not performing her side of the contractual obligation (the so-called use of competition law “as a shield”). The action B, for instance, may consist of not respecting a product exclusivity clause or territorial limitation clause and starting to deal with the upstream competitors of Player I.39 After Player II decides to breach the contract (action B), at the moment of time T2 Player I has another set of two actions: the action S which means to sue Player II for breaching the vertical agreement; or, the action B which means to tolerate the previous breach of Player II. Before moving to the description of the parties preferences and payoffs, we 36

Moreover, this is in line with our previous assumption that at the moment T0 Player II is not aware that the contract is anticompetitive and that she has specific rights under the private enforcement of competition law. 37 This action will cover also the case that he decides to include an arbitration clause provided that the dispute is declared non-arbitrable by the adjudicator. See the previous Chapter for details. 38 See Section 2.1.2, above. 39 This action may be preferable in case the contract was partially performed by the parties.

115

3. THE JURISDICTIONAL ISSUES OF ANTITRUST ARBITRATION UNDER THE LENS OF GAME THEORY

can clarify the structure of this game providing an extensive form representation. Player I C−

C0

Player II S

{...;...}

Player II S

B

Player I S

{...;...}

{...;...}

B

B

Player I S

{...;...}

{...;...}

B

{...;...}

In order to complete the game description it is necessary to assign the parties’ payoffs for each strategy. As mentioned before, the implicit assumption of this game is that the parties have signed an anticompetitive vertical agreement. The results of the legal analysis of the previous Chapter suggest that under an anticompetitive arbitration clause it will be impossible or extremely difficult for the damaged party to obtain compensation. This means that under the C − branch of the three Player I ’s payoff for the strategy (P I(C − ); P II(S)) will be 0, because under that anticompetitive arbitration clause he would not be condemned to pay any damages. Meanwhile, Player II ’s payoff for the suing action S under the C − branch (an anticompetitive arbitration clause) is negative −1 because she does not receive any damages and has to pay the enforcement costs.40 When Player II decides to breach the contract B under the C − branch, in practice she is appropriating the damages. When Player I sues S under the C − branch, the arbitral tribunal will condemn Player II ’s breach and force her to comply with 40

In this Chapter, from this game onwards we use also negative payoffs to order the parties preferences. The objective of this notation is to reduce the chance of confusion between this game payoffs structure and the previous one. The two games are not connected and there should be no comparison between payoffs.

116

the contract, as if the contract was never breached. The payoffs for the strategy (P I(C − ); P II(B); P I(S)) are 0 for Player I (who does not pay any damages and is in the same position as if the contract was duly performed) and −1 for Player II because she does not receive any damages and has to pay the enforcement costs. The last possible strategy under the C − branch is (P I(C − ); P II(B); P I(B)). Player I ’s breach action B means that he is tolerating Player II ’s appropriation of damages. Therefore the payoffs for this last strategy are −1 for Player I (in practice he is paying the damages) and 1 for Player II (she has appropriated the amount of damages breaching the contract and acting as if the contract was pro-competitive). Shifting now our attention on what happens under the C 0 branch, as clarified above, this action means that Player I either decides to remove the arbitration clause from the contract or his legal system does not allow the arbitrability of antitrust disputes. This means that, the antitrust private enforcement dispute will be decided by a national court.41 If Player II sues for damages undertaking the action S, she receives the damages and her payoff is 1; whereas Player I ’s payoff for the strategy (P I(C 0 ); P II(S)) is −2 because he has to pay the damages and the litigation costs.42 The same outcome and payoff assignment holds true also in case Player II decides to breach the contract, following the B action. Player I can choose to sue Player II (action S) in T2 , losing the case and having to pay damages plus litigation costs to Player II. Alternatively, in T2 Player I can decide to tolerate the breach of the contract (action B) and lose only the damages (not the litigation costs) appropriated by Player II when she breaches in T1 . The resulting final payoff assignation in this latter case is for Player II equal to 1 and for Player I equal to −1. This strategy (P I(C 0 ); P II(B); P I(B)) results basically in the same outcome that would have occurred in case the vertical agreement was pro-competitive in the first place. As a matter of fact, if 41

The reader should remember that this game assumes that the vertical agreement is anticompetitive. This assumption is necessary to analyse the effects of different procedural rules, instead of discussing the substantive antitrust rules. 42 We implicitly adopted the English Rule of litigation costs allocation. Because this work focuses on the EU scenario where the American Rule of litigation costs allocation is practically absent. However it can be easily demonstrated that the analysis does not change even employing the American Rule.

117

3. THE JURISDICTIONAL ISSUES OF ANTITRUST ARBITRATION UNDER THE LENS OF GAME THEORY

Player I tolerates the breach of the anticompetitive vertical restraints it is the same as saying that those vertical restraints were not put in place and that the contract is ultimately pro-competitive. The figure below is the extensive form representation of the game that we have just described, including the payoff assignation. Using backward induction, it is possible to determine the outcome of the game and to understand whether the lawmaker was correctly afraid that providing the parties with the option to include an arbitration clause would produce undesirable results. If this is the case, it means that the procedural rule denying the arbitrability of antitrust disputes is capable to enhance the effectiveness of a private enforcement system. Player I C−

C0

Player II S

{-2;1}

Player II S

B

Player I S

{-2;1}

{0;-1}

B

B

Player I S

{-1;1}

{0;-1}

B

{-1;1}

Judging from the extensive game above, it seems that the lawmaker has correctly assumed that limiting antitrust arbitrability is the best choice. In case arbitrability is allowed, Player I has a dominant strategy (C − ; S). Independently of the counterparty’s action, when allowed, he will include into the contract an anticompetitive arbitration clause that limits the effectiveness of antitrust private enforcement in arbitration. The result is that under this clause, it is impossible or extremely difficult for Player II to enforce her competition law rights. In prac-

118

tice, the parties will return to the equilibrium described in the first strategic form payoff matrix described in the previous Section. However, the problem with this theory and the results reached is not on the C branch, but underneath the C 0 branch of the game. In fact, the entire game description of preferences and payoffs does not take into consideration what happens in the context of international commerce. There are objective difficulties connected to the circulation of national courts’ decisions across different legal systems. Conversely, as clarified in the introductory Chapter, it is considerably easier to enforce international arbitral awards through the New York Convention than to enforce national court’s decisions.43 Therefore limiting the arbitrability of competition law disputes reduces also the effectiveness of private antitrust enforcement in international disputes. −

In order to clarify this point we need to describe the parties’ preferences and payoffs when they enforce antitrust rights based on international transactions in front of a national court. This means that the game description of the C − branch is not affected by any change. Regarding the C 0 branch, Player II receives a positive payoff only in case she decides to breach the contract (action B) in T1 . As explained in the legal analysis, if either Player II or Player I choose to sue (action S), the national court’s decision on antitrust disputes is not easily enforceable abroad. The payoff for Player II ’s S action has to be negative −1 because his national court’s decision is limitedly enforceable abroad and results in no compensation for damages but only losses in litigation costs. However, when Player II can decide to breach the vertical restraints, action B, her payoff will be positive and equal to 1. In practice, when she breaches the contract, Player II withholds the amount of damages. This result is independent of the action undertaken by Player I. In fact also Player I faces the problem of enforcing a national court decision abroad. When he chooses to sue (action S) in T1 , Player I has to pay the litigation costs in front of his national court, which can hardly oblige a foreign Player II to comply with the contract.44 Therefore we can as43 44

See Section 1.1.2, above. In the legal analysis of the previous Chapter we explain both: why Player I has to sue in

119

3. THE JURISDICTIONAL ISSUES OF ANTITRUST ARBITRATION UNDER THE LENS OF GAME THEORY

sign Player I ’s payoff for (P I(C 0 ); P II(B); P I(S)) a double negative value of −2. In fact, when Player I action is to breach the contract, B, his payoff for (P I(C 0 ); P II(B); P I(B)) is “less negative”, so −1, because he is accepting the breach but at least he does not have to pay litigation costs. We provide below the extensive form representation of this game that takes into account the effectiveness of C 0 in the context of international commerce. Player I C−

C0

Player II S

{0;-1}

Player II S

B

Player I S

{-2;1}

{0;-1}

B

B

Player I S

{-1;1}

{0;-1}

B

{-1;1}

Also in this game tree, if Player I is allowed to choose, his dominant strategy is still to choose the branch C − . However this game underlines that in the context of international commerce, even when the lawmaker limits antitrust arbitrability,45 the outcome of the game is incompatible with the one desired and described in the previous game. In light of our legal analysis, when there is a competition law breach, the procedural rules should make Player II indifferent between suing the counterparty (use of competition law “as a sword”) or breaching the contract (use of competition law “as a shield”). This choice should be determined only by external (non procedural) variables. The present game theory description shows front of his national court and why we assume that the court decision is hardly enforceable in a different legal system. 45 In other words, when the lawmaker employs a procedural rule forcing Player I to choose the action C 0 .

120

that under the C 0 branch Player II has a dominant strategy, to breach (action B). This means that the effectiveness of the private antitrust enforcement system is limited to Player II ’s action breach B. In other words, Player II can only use his competition law based rights “as a shield” and not “as a sword”.46 Therefore, the present game shows that limitations to arbitrability aiming to address the problems set by the C − branch impair the effectiveness of antitrust private enforcement in international commerce. Moreover, this solution adversely affects international commerce. When Player II is dealing with an international Player I, she will have to internalise the costs and the risks of signing such contract with this Player I in comparison to signing the same contract with a national Player I. In other words, this procedural rule creates an additional costs for international business transactions. On the contrary, we will show below that a procedural rule that supports the wide arbitrability of antitrust disputes paired with a pro-competitive arbitration clause can solve this problem. In fact, international arbitration is not per se incompatible with competition law. A pro-competitive arbitration clause could provide an outcome that increases the effectiveness of antitrust private enforcement in international commerce. Nevertheless, the risk of having in the market vertical agreements that include an anticompetitive arbitration clause C − persists. However, this risk should be tackled with another more appropriate tool by the lawmaker. In order to clarify this argument, we discuss the same extensive game without changing the C 0 branch but instead of comparing it with the anticompetitive arbitration clause resulting from Player I ’s C − action we assume there is a C + action that includes a pro-competitive arbitration clause in the vertical agreement. In other words, we assume that the legal system has a way either to prevent the inclusion of anticompetitive arbitration clauses into the contract or to convert an anticompetitive clause in a pro-competitive one. By definition,47 a pro-competitive arbitration clause provides a wide protection for the damaged party, in our example Player II, also in the context of international transactions. In our game description, this means that independently of the par46 47

See previous Chapter, above. See Section 1.1.2 and Section 2.1.1, above.

121

3. THE JURISDICTIONAL ISSUES OF ANTITRUST ARBITRATION UNDER THE LENS OF GAME THEORY

ties’ actions under the C + branch, if there is a competition law breach, Player II receives compensatory damages and her payoff is 1. As per Player I, under the C + branch, if on the parties chooses to sue (action S), Player I have to pay damages and legal costs receiving a negative payoff −2. Without litigation costs his negative payoff for (P I(C + ); P II(B); P I(B)) is only −1. Player I C0

C+

Player II S

{0;-1}

Player II S

B

Player I S

{-2;1}

{-2;1}

B

B

Player I S

{-1;1}

{-2;1}

B

{-1;1}

Solving with backward induction the game tree above, both Player II and Player I appear to be indifferent between the C 0 action and the C + one. If this was always the case, it would be the best scenario. Antitrust arbitration would not aversely effect private enforcement. Private parties’ decision whether or not to include an arbitration clause into the contract will not be affected by their strategic behaviour on competition law private enforcement procedural rules. However, technically in this game there is a weak dominance of Player I ’s action C 0 . Under the C 0 branch, Player II is paid damages only if she chooses to breach the contract, (action B). Therefore, if Player I has to choose between no antitrust arbitration clause (C 0 ) and a pro-competitive arbitration clause (C + ), Player I would prefer to choose the former. That is because under that strategy he is never worse-off but sometimes he may be better-off. This outcome is reflected in the real world cases. The real world Player I has the incentive to avoid a fulling working

122

pro-competitive arbitration clause that allows Player II to freely choose between suing for damages and breaching the contract. The legal analysis of the previous Chapter addresses cases falling under this scenario. This is when Player I tries to exploit the bilateral benefits of arbitration as an international ADR tool excluding from those ADR benefits disputes related to competition law. Legal scholars have studied these arbitration clauses that only exclude competition law from their scope and that for this reason can be considered anticompetitive.48 In fact, procompetitive arbitration clauses have the potential to overcome the shortcomings of national courts in the context of international commerce, allowing Player II to use competition law not only “as a shield” but also “as a sword”.49 Therefore, the counter-intuitive result of the game theory analysis of this Paragraph tells us that procedural rules that limit the private enforcement of competition law to national courts can make an antitrust private enforcement system less effective.

3.3

Should Competition Law Opt-in or Opt-out of an Arbitration Clause?

From this Chapter emerges that a clear-cut distinction between an anticompetitive arbitration clause and a pro-competitive one is fundamental. However it is not always easy for the adjudicator to evaluate whether an arbitration clause is valid or not. As described in the legal analysis, there are certain elements signalling to the adjudicator that a specific arbitration clause was drafted with the explicit purpose to exclude or limit the private enforcement of competition law. The rest of this Chapter deals with procedural rules that aim to guarantee that anticompetitive clauses are not enforced by an adjudicator. The goal of this Section is to show some adverse effects of a recent judge-made legal rule on antitrust arbitration that has been described in the legal analysis of 48

See Section 2.1.2.1 above. This can prove crucial in some real world cases, such as anticompetitive contracts that have been drafted in such a way to make for Player II the only viable action the sue strategy S. For instance, this is the case for lock-in practices like an anticompetitive contract providing for upfront payments by Player II. If Player II has already paid upfront she cannot breach the contract anymore. 49

123

3. THE JURISDICTIONAL ISSUES OF ANTITRUST ARBITRATION UNDER THE LENS OF GAME THEORY

the previous Chapter.50 This rule requires that private parties explicitly mention competition law disputes in the scope of an arbitration clause, hence we call it opt-in rule. This requirement can make it more difficult for the adjudicator to assess whether he/she is facing a valid antitrust arbitration clause or an invalid one. Under the opposite rule that we can call opt-out rule, when the parties do not explicitly mention competition law in the arbitration clause, antitrust disputes are still considered included in the scope of the clause. If they want to exclude antitrust disputes from the arbitration clause, the parties need to explicitly state so in the arbitration clause. Under the opt-in rule, the opposite is true. When the parties do not mention competition law, it is considered that antitrust disputes are not included into the arbitration clause. Whereas, only if the arbitration clause explicitly refers to competition law disputes, arbitrators will have jurisdiction to assess the matter. When arbitrators have to evaluate whether an arbitration clause violates competition law, they will have to refer to the wording and the design of the arbitration clause itself. In light of the stronger signalling power of a positive action in comparison to a negative action, our argument is that an opt-out rule has a stronger signalling power of anticompetitive clauses in comparison to an opt-in rule. The fundamental assumption in this part of the analysis is that, in light of the results of the previous Section, arbitration is better suited for the private antitrust enforcement in international transactions than national courts. Therefore, when international business parties are not satisfied with the efficiency of national courts they will include an arbitration clause that covers all future disputes based on their contract. It is more difficult for Player I to justify the rationale behind the explicit exclusion of competition law, under an opt-out rule. This action provides a strong signal that can be read by arbitrators as an attempt to obstacle the effective private enforcement of competition law. Therefore arbitrators can better evaluate the invalidity of an arbitration clause. 50

See Section 2.1.2.2

124

On the contrary, under an opt-in rule, the arbitrators cannot give the same strong signalling power to Player I’s silence. This legal rule somehow justifies and legitimises the exclusion of competition law through silence. The explicit inclusion of competition law in the arbitration clause does not add any signalling power to the wording of the clause itself, it just complies with the efficiency of the one-stop principle in litigation, making it harder for either arbitrators or judges to discover anticompetitive arbitration clauses. To summarise, when national courts are less efficient in solving antitrust private enforcement disputes than international arbitrators (this is the case in the context of international business transactions), an opt-out rule is preferable to an opt-in rule, given its stronger signalling power of anticompetitive arbitration clauses.

3.4

Not all Arbitration Clauses are Alike

Instead of limiting antitrust arbitrabiliy the lawmaker can introduce a procedural rule that keeps anticompetitive clauses out of the legal system.51 This result can be reached only if legal systems impose some form of control on arbitration clauses negotiated by private parties. This Section deals with a control mechanism that has been designed to scrutinise arbitration clauses and/or arbitral awards. According to the kompetenz-kompetenz principle, arbitral tribunals are supposed to autonomously determine ex ante the validity of an arbitration clause. However, in light of the fact that international arbitrators are not an expression of the State judiciary power, there is also an ex post second-look review by national judges. The latter is performed when the award requires the State’s ius imperium, namely during the phase of review/recognition/enforcement or the award (the so-called second-look review ). However, as clarified in the legal analysis, adjudicators from different jurisdictions may adopt different choice of law rules to determine the law applicable to determine the validity of an arbitral award. This leads to the application of different sets of procedural and substantive provisions, potentially 51

See Section 2.1.2, above

125

3. THE JURISDICTIONAL ISSUES OF ANTITRUST ARBITRATION UNDER THE LENS OF GAME THEORY

capable to reach opposite outcomes when evaluating the validity of either an arbitration clause or an arbitral award.52 The game theory analysis of this Section deals with the abstract functioning of the kompetenz-kompetenz principle and the second-look review used to censure arbitral awards enforcing invalid anticompetitive clauses. This analysis clarifies the shortcomings of the procedural rules currently in place, while the next Section proposes an alternative solution to address this issue. In all the games described above, each time one of the players was following the strategy S, thus suing his/her counterparty, we just saw the outcome of the dispute with the respective payoff allocation. At this point, the analysis deals with the dispute-resolution sub-game and the related procedural issues. Under an arbitration clause, when a player sues its counterparty (action S), a sub-game begins that has a variable number of other players acting at different moments of time. They all have their information sets, preferences, and action sets which lead to specific strategies and to a specific outcome of the game. After one of the parties files an action for damages, the arbitration proceeding begins and the first new player to act is the Arbitrator. Under the kompetenzkompetenz principle an arbitral tribunal has to solve the jurisdictional issues extensively discussed in the legal analysis of the previous Chapter. In brief, the arbitrators have to determine the validity of an arbitration clause as well as the law applicable to the merits of the case. The decision on both these issues can define the very outcome of a dispute and is strictly related to the answer given to the choice of law rule question.53 The link between these procedural rules need to be simplified so that we can discuss them using game theory analysis. Hence, in light of the legal analysis of the previous Chapter, we can link the Arbitrator ’s decision on the choice of law rule directly to his decision on the validity of the arbitration clause and also to the validity of the vertical restraints.54 Therefore, 52

See Section 2.2 above. See from Section 2.2.2 onwards. 54 As clarified in the legal analysis the choice of law rule affects the validity of both the arbitration clause (determining the applicable procedural law) and the vertical restraint clauses (determining the applicable law to the merits of the case). For simplicity, throughout this 53

126

in this game description, the actions that the Arbitrator have to choose are described as: enforce E, modify M or void V the arbitration clause. Whereas the first action E is self explanatory, in the last two the arbitrator has to decide whether the arbitration clause is, or it is not, totally or partially invalid and anticompetitive. The Arbitrator ’s goal is to produce a valid and enforceable award capable to provide the parties with a decision on the dispute, according to the procedural and substantive legal rules chosen by them. This will increase his reputation in the international arbitration community, therefore increasing his future revenues. Hence, the Arbitrator receives a positive payoff 1 (increasing his reputation) either if the parties are capable to enforce his award (a national court does not set aside the award) or if his evaluation of the arbitration clause is confirmed by the national court in the second-look review.55 In fact, when one of the parties does not agree with the Arbitrator ’s assessment of the clause that party can challenge his evaluation in front of a national judge. Before moving forward in the analysis, a further description of the game, its players and their preferences is necessary. First of all, a simplification of the game can be useful. If the Arbitrator decides to void (action V ) the arbitration clause, the game will not have any new player; it will be substantially identical to the scenario of the games discussed in the previous Sections,56 namely when the only action available for Player I is C 0 and the dispute is solved through ordinary litigation. Therefore, in order to complete the description of the V branch of the game is only necessary to assign the Arbitrator ’s payoffsl since the parties’ preferences and payoffs do not change. As explained before, when Player II sues in front of his national court (action S), the decision is not enforceable; hence Arbitrator ’s payoff for the strategy (A(V ); P I(S)) is 0 (the parties cannot solve the dispute and the Arbitrator ’s decision is useless to them). When Player II breaches the contract and appropriates the value of the damages (action B), Section we mention only the validity of arbitration clauses. Nevertheless the reader should remember that this analysis can be expanded also to the validity of specific vertical restraint clauses. 55 Obviously the Arbitrator receives his positive payoff 1 also when the parties do not challenge the award 56 See Section 3.2, above.

127

3. THE JURISDICTIONAL ISSUES OF ANTITRUST ARBITRATION UNDER THE LENS OF GAME THEORY

Player I can sue her in front of his national court but this decision is not enforced by Player II ’s national court that agrees with the Arbitrator ’s decision.57 However, since the Arbitrator foresaw this outcome taking the action V his payoff for the strategy (A(V ); P II(B); P I(S)) is 1. Obviously if no party challenges the arbitral award and just breaches the vertical agreement the Arbitrator ’s payoff for the strategy (A(V ); P II(B); P I(B)) is still 1. Below is an extensive form representation of the game, limited to the V branch:

E

{...}

Arbitrator

M

{A(1);PII(1);PI(-1)} B

{...}

Player I

V

B

S

{A(1);PII(1);PI(-2)}

Player II

S {A(0);PII(-1);PI=0}

We can make another useful simplification in this game. In fact, in case the Arbitrator decides to enforce the arbitration clause (action E) or to modify it (action M ) in such a manner that it can be considered pro-competitive, the next move will be played respectively by Player II and by Player I. Each one of them is respectively the losing party in the Arbitrator’s decision about the validity of an arbitration clause. We assumed in Section 3.2 that Player I drafts the arbitra57

The reason why Player II ’s national court agrees with this decision can be understood in light of the legal analysis above. We can think about the Coca-cola example. If an arbitral tribunal deems the arbitration clause anticompetitive and sets it aside, Coca-cola can sue its retailer in the US, but it has still to enforce the decision in the EU. As explained above, in this example, the EU national court agrees with the decision of the arbitral tirbunal.

128

tion clause with the intent to avoid private antitrust enforcement in arbitration. This means that if the Arbitrator enforces the award (action E) he achieves his goal. Conversely if the Arbitrator modifies the arbitration clause to make it procompetitive (action M ) Player II can claim his competition law rights and the losing party is Player I. This symmetry in the two players positions is reflected in their preferences and payoffs. In other words the two branches E and M are substantially symmetrical and we can omit describing the M branch. We can focus, instead, on discussing and understanding the E branch because the other has a symmetrical outcome and will follow the same reasoning.58 After the Arbitrator decides to enforce the clause, action E, Player II can either sue (action S), challenging the award in front of the national court of the arbitration seat, namely a player that we call Seat Judge, or to breach (action B) the award disregarding the Arbitrator ’s decision. When the latter strategy is employed, Player I has to decide whether to tolerate the breach of the arbitral award (action B) (so that the award will not be enforced and each party will bear his/her own legal costs) or to sue (action S) seeking the enforcement of the award in front of a national court. The national court that has to decide on the award recognition/enforcement is located in the legal system of Player II ; hence we call this new player Judge II. As we have extensively explained in the legal analysis, our research project focuses on antitrust arbitration issues within an international context. This means that we focus on disputes between international parties belonging to legal systems that have conflicting antitrust regimes. In this context, the party that wants to enforce the arbitral award has to start legal proceedings in front of the national court of its counterparty. This means that Player I has to enforce the award against Player II in front of Judge II. Respectively, under the symmetrical M branch that we decided to omit, Player II has to enforce the award against Player I in front of Judge I.59 Moving back to the description of Player II ’s action set, Player II can decide that instead of breaching the award (action B) she wants to sue (action S) challenging the 58

As we will see below, in the M branch of the game Player I takes the place of Player II and Judge I takes the place of Judge II. 59 See Section 2.2, above. Obviously the parties are facing the same choice also after the action of the Seat Judge.

129

3. THE JURISDICTIONAL ISSUES OF ANTITRUST ARBITRATION UNDER THE LENS OF GAME THEORY

validity of the award in front of the Seat Judge. At this point the Seat Judge has to decide whether to set aside the award, voiding it (action V ), or to confirm the award enforcing it (action E). In the latter case, Player I is in the same position as after Player II decided to breach (action B) the arbitral award in the first place. Therefore, Player I has to decide whether or not to seek the award recognition/enforcement in front of Judge II. At this point, the game description can focus on the actions of all the judges involved: Seat Judge, Judge II and Judge I (who acts instead of Judge II under the M branch). The results of legal analysis tell us that if a national court applies the same choice of law rule that arbitrators have applied, it is more likely to enforce the arbitral award. Otherwise the national court’s assessment of the jurisdictional issues and the merit of the case is incompatible with the arbitrators’ decision; hence the national court will set aside the arbitral award. This means that, in our game theory description, all national courts have to choose between two available actions: void the clause and the award (action V ), or enforce the clause and the award (action E). For the sake of clarity in theory the judges have also the possibility to modify (action M ) the arbitration clause.60 However this last action implies that the award must be set aside. Hence, a new arbitration proceeding will have to begin, making the whole game start anew.61 Therefore, the outcome of the action M is substantially identical to the one of the action void V . This means that, for the sake of simplicity, national courts’ action M can be omitted in this game.62 In order to facilitate the understanding of this game, the structure of the E branch (without payoffs) was added below to the previous extensive form representation of the game including the V branch (with payoffs). We have also removed the M branch because of its redundancy. 60

This last option implies that modifying the arbitration clause the national court will modify the elements that are relevant to determine the choice of law rule applicable. 61 The reason for this is the fact that in general a national court cannot solve a dispute and decide the merits of a case if the parties have agreed that a disputes should be solved in arbitration and not in litigation. 62 Also the costs allocation between the parties will be the same. When a judge decides that a clause is anticompetitive and modifies it M , as well as when he sets the clause aside and considers the clause void V , the losing party is Player I. In both cases, he is the one who has to pay the costs of the arbitration proceedings.

130

{A();PII();SJ();PI()}

V {A();PII();SJ();PI();JII()} V

Seat Judge

E

Judge II

S

S Player I

B

E

{A();PII();SJ();PI();JII()}

{A();PII();SJ();PI()}

Player II

{A();PII();PI();JII()} V

B

Judge II

S

E

E

{A();PII();PI();JII()}

Player I

B {A();PII();PI()} Arbitrator {A(1);PII(1);PI(-1)}

B Player I

V

B

S {A(1);PII(1);PI(-2)}

Player II

S {A(0);PII(-1);PI(0)}

131

3. THE JURISDICTIONAL ISSUES OF ANTITRUST ARBITRATION UNDER THE LENS OF GAME THEORY

In order to complete the description of this game we need to define each player preferences and payoffs. At the beginning of this Section, we have already explained the Arbitrator ’s preferences and how his payoffs are assigned. We have also extensively discussed the parties preference and payoffs allocations. However it is worth re-framing the parties’ payoff assignation in the context of this arbitration sub-game under the E branch. Player I receives a positive payoff 1 if the arbitral award is enforced by Judge II (action E). He receives a negative payoff −1 if the award is set aside by a national court (action V of a national court, including the Seat Judge). Obviously, the opposite happens for Player II, she receives a negative payoff −1 if the arbitral award is enforced by Judge II (action E) and a positive payoff 1 if the award is set aside by a national court (action V of a national court, including the Seat Judge). Both players’ payoff for the strategy (A(E); P II(B); P I(B)) is 0 because this strategy means that the parties do not give any value to the arbitral award. Obviously if the arbitral award is useless for the parties also the Arbitrator’s payoff for the strategy (A(E); P II(B); P I(B)) is 0. Finally, Player I ’s payoff for the strategy (A(E); P II(S); SJ(E); P I(B)) is also 0 because this strategy means that after the the Seat Judge confirms enforces the award (action E), Player I renounces to enforce it in front of Player II ’s national court. In this last scenario Player II still receives a negative payoff because she lost in front of the Seat Judge, she does not receive any compensation in damages and she had also to pay the litigation costs of the Seat Judge. Regarding the preferences of each judge, in light of the legal analysis,63 this game assumes that the arbitration clause is anticompetitive and invalid according to the legal system of Player II and Judge II, while it is valid in the legal system of Player I and Judge I as well as in that of the Seat Judge. The reason for the latter (the equivalence of preferences between Judge I and Seat Judge) is based on the fact that the arbitration seat is chosen by Player I in the arbitration clause.64 All these assumptions reflect the issues discussed in the legal analysis above. From these assumptions we can imply that Judge II gains a positive rep63

See from Section 2.2.4 onwards. See Section 2.2.2, above. Where it is described why legal scholars do not consider the lex arbitri a suitable choice of law rule. 64

132

utation payoff 1 when he sets aside (action V ) the award that applies the original arbitration clause, and he receives 0 otherwise. Symmetrically, under the omitted M branch of the game Judge I gains a positive reputation payoff 1 when he sets aside (action V ) the award that has modified the arbitration clause, and he will receive 0 otherwise. As per the Seat Judge, Player I has the incentive to draft the arbitration clause choosing the arbitration seat in such a way that the Seat Judge belongs to a jurisdiction tolerant with clauses like the one that he drafted. In other words Player I will position the arbitration seat in the legal system that provides the higher protection for the original arbitration clause. Therefore we can assume the Seat Judge will gain a positive reputation payoff 1 confirming the validity E of the award that applies the original arbitration clause, and he will receive 0 otherwise. Likewise, in the omitted M branch the Seat Judge will gain a positive reputation payoff 1 setting aside V of the award that modified the original arbitration clause, and he will receive 0 otherwise. Even if we tried to simplify the game description as much as possible, this game is fairly complex to explain in words. The extensive form representation of the game, below, should provide some help.

133

3. THE JURISDICTIONAL ISSUES OF ANTITRUST ARBITRATION UNDER THE LENS OF GAME THEORY

{A(0);PII(0);SJ(0);PI(-1)}

V {A(0);PII(1);SJ(1);PI(-1);JII(1)} V

Seat Judge E

Judge II

S

S

E

Player I {A(1);PII(-1);SJ(1);PI(1);JII(0)} B {A(1);PII(-1);SJ(1);PI(0)}

Player II

{A(0);PII(1);PI(-1);JII(1)} V

B

Judge II

S

E

E

{A(1);PII(-1);PI(1);JII(0)}

Player I

B {A(0);PII(0);PI(0)} Arbitrator {A(1);PII(1);PI(-1)}

B V

Player I

B

S {A(1);PII(1);PI(-2)}

Player II

S {A(0);PII(-1);PI=0}

134

Using backward induction on the extensive representation of the game above we can solve it and find the outcome of the game. The Arbitrator does not have the incentive to choose the action E (nor the symmetrical action M that we have omitted) because the arbitral award will not be enforced, both parties will breach following the strategy (A(E); P II(B); P I(B)) and the Arbitrator will lose a potential reputation payoff of 1. In fact, under the E branch of the game, Judge II will always set aside the award (action V ), which gives to Player I a negative payoff −1 for either the strategy (A(E); P II(S); SJ(E); P I(S); JII(V )) and the strategy (A(E); P II(B); P I(S); JII(V )). Hence Player I has a dominant strategy breach (action B) that gives him a payoff of 0 either for the strategy (A(E); P II(S); SJ(E); P I(B)) and for (A(E); P II(B); P I(B)). Also Player II ’s dominant strategy under the E branch is to breach (action B) receiving a payoff 0 for the strategy (A(E); P II(B); P I(B)). When Player II sues (action S), the Seat Judge will enforce the award, she has to pay the litigation costs of the Seat Judge and when Player I breaches (action B), Player I ’s payoff for (A(E); P II(S); SJ(E)P I(B)) is equal to −1. Therefore the only action that provides the Arbitrator with a positive payoff is the action V (void the arbitration clause). However, as mentioned above, this option is substantially identical to a national antitrust private enforcement system. It was called above the C 0 action of Player I, and it implies all the shortcomings that we have analysed in the previous Sections. Obviously we do not have any empirical evidence of arbitral awards that have not been enforced. However the legal analysis of the previous Chapter mentions the fact that, when it is possible (namely when the parties do not explicitly raise the issue), arbitrators try to avoid dealing with competition law issues. This arbitral tribunals’ attitude could be explained as a reflection of the results of our game theory analysis. To summarise, the currently in force second-look review procedure along with the lack of uniformity in the choice of law rule do not guarantee that private parties employ only pro-competitive arbitration clauses. On the contrary, our game theory description suggests that these procedural rules only impair the effectiveness of private antitrust enforcement in arbitration. There could be alternative

135

3. THE JURISDICTIONAL ISSUES OF ANTITRUST ARBITRATION UNDER THE LENS OF GAME THEORY

solutions more suitable to address the problem of anticompetitive arbitral awards (arbitral awards based on anticompetitive clauses). The solution proposed more often, in order to correct this problem, is to change the procedural rules governing the second-look review of the award. This could be a viable option to address the problem. However the legal literature has debated on the second-look extensively, either repeating the same arguments or simply clashing with the diverging solutions adopted by the case-law of different jurisdictions.65 Maybe it is time to accept that international uniformity with respect to the second-look review is never going to succeed.66 The next Section suggests an alternative solution based on the pursuit of uniformity among the choice of law rules adopted by the adjudicators. The solution below can be applied both to the choice of law rule used for deciding the law applicable either to the jurisdiction or to the merits.

3.5

A Possible Solution to the Conflict of Law Issues in Antitrust Arbitration

In the previous Sections, the discussion focused mainly on procedural rules aiming to verify the validity of both arbitration clauses and arbitral awards. The legal analysis of the previous Chapter clarifies that in order to determine the conformity of these two elements with competition law, the essential prerequisite is to identify the law applicable both to the jurisdictional issues (validity of the arbitration clause) and to the merits of the case (arbitrators’ assessment of the vertical restraints validity). Therefore, the definition of a uniform choice of law rule is the cornerstone of international antitrust disputes. In fact, the arbitrators’ choice of law rule in antitrust disputes is a concern for national courts and is the central topic of this Paragraph. Competition law is a set of mandatory rules that are: fundamental for the functioning of a legal system; subject to extraterritorial application; and that cannot be violated without also violating the public order of a state. Moreover, in a context of international commerce, the 65

Radicati di Brozolo, L., Chapter 22 in Blanke and Landolt [2010]. This uniformity could guarantee a higher degree of certainty to the private parties’ initial expectations about the validity of an arbitration clause. However such a uniformity may not even be necessary if there is another way to reach the same result. 66

136

latter characteristic, namely the extraterritorial applicability, makes this set of mandatory norms particularly inclined to generate a conflict of law issue with other legal systems. As explained in the legal analysis above, the conflict of law issue becomes extremely important when either procedural or substantial rules are inconsistent among legal systems. In fact, the lack of uniformity means that a decision on the choice of law rule will affect the very outcome of the dispute.67 In order to clarify the problem underneath the choice of law rule, it could be useful to re-frame our Coca-Cola example from Section 2.2.1 in light of the game theory settings of this Chapter. Coca-Cola is our Player I and Player II is the European distributor. Coca-Cola puts an arbitration clause and certain vertical restraints into the distribution agreement (for instance territorial segmentation). We assume that these vertical restraints do not violate the US antitrust law, while they are considered anticompetitive under the EU competition law due to their incompatibility with the internal market. Each party can have certain expectations about the antitrust law applicable to the contract, either US law or EU law. In any case, the parties may have either the same or different expectations about the law applicable to the vertical agreement. The parties will perform the contractual obligations that do not breach what they believe to be the applicable antitrust law. When Player I considers applicable the US antitrust law, if Player II breaches the restraints, Player I can sue her in front of an Arbitrator claiming that she has breached a perfectly valid vertical restraint under the applicable US antitrust law. Symmetrically, when Player II considers applicable the EU competition law, she can defend herself claiming that there is no breach, since the restraints are void under the EU competition law, which she deems to be applicable.68 However, the parties’ expectations and believes may be disregarded either by the Arbitrator or by the national court deciding on the matter. The reason for this is the nature of competition law as a set of mandatory norms with 67

As clarified by Section 3.1 above, vertical restraints are one of those topics where there are still relevant inconsistencies in the approach of different legal system. Moreover, also the circulation of the adjudicator’s decision is jeopardised by the lack of a consistent rule to solve the conflict of law question. 68 For the sake of completeness, we should not forget that Player II can also claim that, in light of the anticompetitive clauses, she is also entitled to receive a compensation for damages.

137

3. THE JURISDICTIONAL ISSUES OF ANTITRUST ARBITRATION UNDER THE LENS OF GAME THEORY

extraterritorial application along with the lack of a uniform choice of law rule.69 In our game theory analysis this issue can be represented including Nature as a player. Nature determines which is the law applicable to the dispute. After Nature has determined the law applicable the order of play requires that the Arbitrator decides whether the law applicable is the US antitrust law (U S action) or the EU competition law (EU action). However, at this point of the game we have to represent the uncertainty of choice of law rule. This uncertainty can be represented providing the Arbitrator with an incomplete information set. In other words, he does not know what is the law applicable according to Nature. Nevertheless, the Arbitrator has to make an action. In our Coca-Cola example, if he decides that the US law is applicable, Coca-Cola will win the arbitration proceeding. The contrary will be true, in case he decides that the EU law is applicable. The problems begin when the losing party, either Coca-Cola or the EU distributor, does not comply with the arbitral award. At that point the winning party has to ask the recognition/enforcement of the award in front of a national court located in the legal system of the losing party. This means that a national court of the losing party has to perform a second-look review of the award. The national court that will be reviewing the award is the US court Judge I if the law applied by the Arbitrator is the EU law (EU action).70 Alternatively, the national court that will be reviewing the award is the EU court Judge II if the law applied by the Arbitrator is the US law (U S action).71

69

See Section 2.2, above. Because in that case the winning party is the EU distributor Player II. 71 Because in that case the winning party is Coca-Cola Player I. 70

138

The structure of this game is represented below in an extensive form with order of play and information sets but (for now) without the parties payoffs: {..;..;..}

EU Judge I

EU

US {..;..;..}

Arbitrator {..;..;..}

U

US

EU

E

Judge II

US {..;..;..}

Nature

{..;..;..}

EU

U S

Judge I

EU

US {..;..;..}

Arbitrator {..;..;..}

US

EU Judge II

US {..;..;..}

139

3. THE JURISDICTIONAL ISSUES OF ANTITRUST ARBITRATION UNDER THE LENS OF GAME THEORY

In order to determine the outcome of this game it is necessary to describe the preferences of all the players involved. Both parties receive a positive payoff 1 if the national court eventually applies the same law that has been determined by Nature. In fact, if this is the case, there is no uncertainty nor inconsistency between their ex ante expectations on the law applicable and the law that will be eventually applied in litigation.72 Since in this game the parties’ preferences are aligned there is no need to express them twice. Both parties have the same payoff that is indicated in extensive form representation of the game as P (1) or P (0). Consistently with the description of the previous games, also here the Arbitrator receives a positive reputation payoff A(1) if his action is confirmed when a national court makes the same action. On the contrary, when the Arbitrator ’s action is in contrast with the action of national court, the arbitral award is set aside, arbitration was useless for the parties and the Arbitrator receives a reputation payoff equal to A(0). The key element that determines the outcome of this game is the choice of law rule that each national court follows when deciding whether to grant extraterritorial application to the antitrust rules of a different legal system. This legal rule affects both the preferences and the payoffs of Judge I, Judge II. As clarified in the legal analysis of the previous Chapter, the conflict of law issue can be solved according to a choice of law rule that requires national courts to apply competition law as “strictly mandatory law”. Under this rule both Judge I and Judge II disregard the action of Nature and receives a positive payoff JI(1) or JII(1) when they apply to the dispute their national antitrust law. Whereas each national court receives a payoff equal to 0 when it applies to the dispute a foreign competition law. Now that the game description is complete we can provide the extensive form representation of the game: 72

It is worth noting that this is in line with the methodology chosen for this research project in the Preface Chapter. This work analysis how different procedural rules affect the parties incentives in litigation. Therefore we need to pay close attention to the parties payoffs even if they are not acting in this game, but Nature is.

140

{P(1);A(1);JI(0)} U E Judge I

EU

US {P(0);A(0);JI(1)}

Arbitrator {P(1);A(0);JII(1)} U E

U

US E

Judge II

US {P(0);A(1);JII(0)}

Nature

{P(0);A(1);JI(0)} EU

U S

Judge I

EU

US {P(1);A(0);JI(1)}

Arbitrator {P(0);A(0);JII(1)} EU

US Judge II

US {P(1);A(1);JII(0)}

141

3. THE JURISDICTIONAL ISSUES OF ANTITRUST ARBITRATION UNDER THE LENS OF GAME THEORY

Provided that there is no outcome in the game where both a national court (either Judge I or Judge II ) and the Arbitrator have a positive payoff, it is clear that this game does not have any outcome in which the arbitral award is enforced and the antitrust arbitration is effective in providing legal certainty for the parties. Therefore, international commerce will be adversely affected by a choice of law rule that sees competition law as “strictly mandatory law”. This scenario resembles the description in the legal analysis of the European and Russian courts’ approach on the conflict of law issues in competition law.73

There is also an alternative choice of law rule, currently in place, that favours international commerce. Under this procedural rule, one of the two national courts follows a different choice of law rule that changes his preferences on law applicable. This approach consists in a less strict interpretation of competition law as a set of mandatory rules. In light of the legal analysis above, this is what could actually happen if our Coca-Cola example was real case study. Such a scenario would resembles the real case-law of European and US national courts.74 Under this approach, the US court, our Judge I, receives a positive payoff (JI(1)) also enforcing of an arbitral award which applies EU competition law. The US legal system considers the EU competition law to be applicable also on US based cases. On the contrary the EU court, Judge II, does not change his preference for a stricter choice of law rule. In other words his payoff structure from the game above does not change. Below is the new extensive form representation of the game:

73 74

See Section 2.2.5, in the previous Chapter. See Section 2.2.4, above.

142

{P(1);A(1);JI(1)}

EU Judge I

EU

US {P(0);A(0);JI(0)}

Arbitrator {P(1);A(0);JII(1)} U E

U

US E

Judge II

US {P(0);A(1);JII(0)}

Nature

{P(0);A(1);JI(1)} U E

U S

Judge I

EU

US {P(1);A(0);JI(0)}

Arbitrator {P(0);A(0);JII(1)} EU

US Judge II

US {P(1);A(1);JII(0)}

Contrarily to the previous game, this game has an equilibrium, the one in the circle above. In fact there is a dominant strategy for the Arbitrator, to apply the EU competition law no matter what is the action of Nature. When also according to Nature the EU law is applicable to the case, under the strategy (Nature(EU);A(EU);JI(EU)) each player receives a positive payoff. Also Judge II receives a positive payoff because his national law has been applied to the dis-

143

3. THE JURISDICTIONAL ISSUES OF ANTITRUST ARBITRATION UNDER THE LENS OF GAME THEORY

pute, satisfying his preference. Therefore, under a more liberal approach on the choice of law rule Judge I has now changed his preferences accepting to enforce the award no matter whether the US law or the EU law is applied by the Arbitrator. While this outcome may not have adverse effects on international commerce, the problem with this solution is systemic. The unique equilibrium of this game leads to the exclusive application of EU law by international arbitration every time there is a conflict of law issue. This means that US law will not be applied even when it should be applied according to what emerges from the action of Nature. As argued in the legal analysis, while this solution could be considered acceptable by the more liberal US legal system in respect with EU competition law, it is not acceptable to other legal systems, such as the Russian one.

The solution to this conflict of law problem already exists and it is currently used by antitrust authorities. However national courts do not apply this choice of law rule. They still retain their discretionary power to set aside correct arbitral awards. The solution that we are referring to has also not been implemented by lawmakers yet. Hence currently there is just a legal gap on this subject matter. The solution is to adopt as a choice of law rule the so-called doctrine of effects rule. Under this doctrine, the law applicable to the dispute is the law of the legal system where the contract has the potential to produce anticompetitive effects. Under this choice of law rule, the parties are able to build certain expectations about the law applicable to the contract and behave accordingly. The arbitrators have to verify which legal system is likely to be affected by the potential anticompetitive effects. Arbitra tribunals have to show explicitly in the award on which basis they decided that the effects are produced in a certain legal system, determining the application of its competition law. The judges performing the “second-look review” have to enforce the well argued arbitral award, without taking on themselves the task of performing a new analysis of effects but just verifying the consistency of the reasoning of the arbitral award. In other words, under the doctrine of effects the parties and the Arbitrator focus on obtaining the information of how has Nature acted in the phase T0 of the game. Therefore this game becomes a perfect information game and each player receives a positive payoff 1 if he/she applies to the dispute the competition law regime that is

144

applicable according to the action of Nature. Such objective choice of law rule will provide the following extensive game representation.

{P(1);A(1);JI(1)}

EU Judge I

EU

US {P(0);A(0);JI(0)}

Arbitrator {P(1);A(0);JII(1)} U E

U

US E

Judge II

US {P(0);A(1);JII(0)}

Nature

{P(0);A(1);JI(0)} U E

U S

Judge I

EU

US {P(1);A(0);JI(1)}

Arbitrator {P(0);A(0);JII(0)} U E

US

Judge II U S {P(1);A(1);JII(1)}

In this last game where procedural the choice of law rule applied follows the doctrine of effects, there are two possible equilibria. When according to Nature

145

3. THE JURISDICTIONAL ISSUES OF ANTITRUST ARBITRATION UNDER THE LENS OF GAME THEORY

the effects of the contract are in the US, the parties expect that the US law is applicable and vice-versa. The Arbitrator can investigate where the antitrust effects are likely to be produced and applies either the US law or the EU law accordingly. The European Judge II or the US Judge I enforces the award applying respectively US law and EU law. If the choice of law rule follows the doctrine of effects each party receives a positive payoff 1. In this way, legal certainty improves and there are favourable effects on international commerce. The main issue with this argument is that the international uniformity of the choice of law rule is linked to a factual analysis based on the objective economic effects of the allegedly anticompetitive conduct. Hence, the key element (or more precisely the requirement) for the proper functioning of this choice of law rule, which is also the main argument against its implementation, is the assumption that the arbitrators are provided with sufficiently strong investigative powers to correctly assess in which legal system are the anticompetitive effects likely to be produced. This issue is the central matter of analysis of the next Chapter. To summarise, we believe that through the use of game theory, this Chapter improves the understanding of the effects that different legal rules have on parties’ preferences and incentives. The analysis provides useful insights for both lawmakers and adjudicators when dealing with the three major jurisdictional issues described in the previous Chapter. The analysis of Chapter allows us: (i) to argue in favour of a more wide and flexible interpretation of the concept of competition law arbitrability; (ii) to claim that procedural rules limiting arbitrability and the expanding the “second-look review” are not the right tool to fight invalid arbitral awards or anticompetitive arbitration clauses; finally, (iii) to show the fallacy behind the current solution to the conflict of law problem due to the extraterritorial application of competition law, and to propose an alternative solution upon which the next Chapter will expand.

146

Chapter 4 Arbitral Tribunals’ Powers to Apply EU Competition Law The economic analysis of the previous Chapter addressed the jurisdictional issues of antitrust arbitration. It advocated for a wide objective and subjective arbitrability assessed by arbitrators and restrictively verified by national judges under the same choice of law rule used by arbitrators and based on the doctrine of effects. This position is grounded on the antitrust arbitration’s potential to guarantee a wider enforcement of competition law rights in international trade. The main obstacles on this route are the legal gap in this field of law and the lack of uniformity among the relevant rules. Although we are not recommending a complete harmonisation of the involved legal rules, the jurisdictional issues emerging in the previous Chapter suggest that the effectiveness of international antitrust arbitration could profit from a uniform choice of law rule. The proposed choice of law rule is based on the doctrine of effects. This doctrine, established by US and EU competition authorities, requires a factual objective analysis of the merits of a case. It migrates the focus of this research project from issues of jurisdiction to issues of merits of a case. In that respect, an arbitral tribunal that needs to decide upon questions of merits must be entitled to the judicial powers necessary to correctly solve the dispute. In other words, when solving disputes based on EU competition law, the arbitral tribunal must

147

4. ARBITRAL TRIBUNALS’ POWERS TO APPLY EU COMPETITION LAW

be entitled to a set of powers that is necessary to comply with the enforcement standard required by the public policy rules. Therefore, the central topic of this Chapter revolves around the analysis of the arbitrators’ powers and the results show that in order to have an effective antitrust arbitration they must comparable to the powers available to national judges.

Legal systems grant to adjudicators certain powers that will help them to solve both factual issues and legal issues. Accordingly, this Chapter is divided in two Sections. On the one hand, regarding factual issues, this work does not cover the investigative powers of arbitral tribunals in comparison to those of national courts. The reason is that in arbitration practice this is usually not a problematic point. Private parties generally comply with their requests. Whereas, when third parties are involved and do not comply, it is widely accepted that national courts will cooperate with arbitrators to obtain the needed evidence.1 What is of interest for this research project is the cooperation mechanism between arbitral tribunals and competition authorities. Antitrust issues on the merits require a fairly complex fact-intensive economic analysis based on market information that only competition authorities can obtain. Therefore, in the European context, the question of the arbitrators’ “powers” to cooperate with the European Commission becomes crucial. This is the central topic of the first Section of this Chapter. On the other hand, with regard to the adjudicator’s powers when facing legal issues, the European Treaties entitle national courts to refer the problem to the CJEU for a preliminary ruling. The second Section of this Chapter deals with the arbitral tribunals’ “power” to request a preliminary ruling under article 267 TFEU. In light of the second-look review of antitrust arbitration, this topic is fundamental. As it will be discussed below, in some circumstances it may be difficult for arbitrators to meet the requirements of the extended second-look review if they cannot refer a case to the CJEU for preliminary ruling.

1

Obviously this is an oversimplification, the investigative powers of arbitral tribunals are an extensive field of studies. However the issues discussed therein are not with the scope of this research project.

148

4.1

Cooperation Between Arbitral Tribunals and the European Commission

Before delving into the legal analysis of the mechanisms through which arbitral tribunals can cooperate with the European commission, this Section resumes one point left open at the end of the previous Chapter. In fact the arbitrators’ power to cooperate with antitrust authorities is of fundamental importance not only to correctly decide on the merits of the case but it could also be the key to solve the conflict of law issue discussed before. We argued in the previous Chapter that the choice of law rule that should harmoniously be applied in international arbitration is based on the doctrine of effects discussed before. However, neither arbitral tribunals nor national courts are the most appropriate subject to perform the fairly complex fact-intensive economic analysis based on market information that can be necessary to apply the doctrine of effects,2 whereas competition authorities have all the necessary tools to perform this analysis. Moreover, allowing adjudicators to cooperate with competition authorities will not only provide more certainty and uniformity into the conflict of law in antitrust arbitration, but it will also provide a mechanism to address public policy related clashes between jurisdictions. To clarify this point it is useful to recall the Gazprom example used in the previous Chapter. Assuming there is a EU competition law breach, the European Commission could coordinate with the Russian competition authority to determine whether the breach produces effects on the EU internal market. The advantage of this scenario over the present one is that in a game theory perspective the two competition authorities will not be playing a one-shot game (as national courts and arbitrators do when they have to solve the conflict of overlapping antitrust regimes) but instead they will be playing a 2

It could be argued that this is the reason why only competition authorities use this tool to cope with the extraterritorial effect of antitrust rules. Meanwhile, national courts and arbitral tribunals struggle with the more traditional choice of law rules discussed in the previous Chapter.

149

4. ARBITRAL TRIBUNALS’ POWERS TO APPLY EU COMPETITION LAW

repeated game. In this context the two competition authorities have the incentive to cooperate. The cooperation could be even taken to the level of agreeing that only part of the anticompetitive conduct affects the internal market. Therefore, under the politically delicate scenario of our Gazprom example, a conciliatory solution could be possible. If only part of the effects is in the EU market then the damaged party will be entitled only to part of the damages. In any case it is not within the scope of this work to discuss and analyse all the advantages and problems emerging from this cooperation mechanism between international competition authorities. Nevertheless this argument clarifies the importance of a cooperation between arbitral tribunals and the European Commission. The next step is to discuss whether such mechanism exists and how it works.3

4.1.1

Regulation 1/2003 and the Cooperation Mechanisms With National Courts

The Modernisation Regulation4 1/2003 establishes, for the first time with hardlaw provisions, the right of national courts to ask assistance from the European Commission. The first paragraph of Article 15(1) of Regulation 1/2003 reads as follows: ‘courts of the Member States may ask the Commission to transmit to them information in its possession or its opinion on questions concerning the application of the Community competition rules.’5 Once established, this general right needs further specification. Hence the Cooperation Notice6 clarifies the procedure for this cooperation mechanism, including the deadlines by which the 3

This cooperation mechanisms may raise concerns about the interference that the competition authorities’ own interests may have with the assessment of the case. These risks can be mitigated by a provision similar to what stated by Article 15 of Regulation 1/2003. This provision states that it is entirely within the discretion of national courts to decide whether to ask for this form of assistance. Moreover on the basis of this provision it has been clarified that the requesting court is not formally obliged to conform its decision to the outcome of this assistance. Namely the parties can convince the adjudicator that the competition authority’s decision has been affected by its own interests. 4 Council Regulation 1/2003 of 16 December 2002 on the implementation of the rules on competition laid down in Articles 81 and 82 of the EC Treaty, OJ L 1, 4.1.2003. 5 Council Regulation 1/2003 of 16 December 2002 on the implementation of the rules on competition laid down in Articles 81 and 82 of the EC Treaty, OJ L 1, 4.1.2003. 6 Commission Notice on Cooperation between the Commission and Courts of the EU Member States in the Application of Articles 81 and 82 EC, OJ 2004 C101/54. This notice is part of the so-called ’Modernisation Package’.

150

Commission must reply.7 In practice national courts can seek two types of assistance:

• documents in possession of the Commission, information about a current proceeding before the Commission. The Cooperation Notice promises that the Commission will respond to such requests within a month;8 and • the Commission’s opinion on economic, factual, and legal matters. The Commission will aim to do so within four months.9 Moreover it is worth noticing that this request is available only if other tools (i.e. the case law of the EU courts and Commission regulations, decisions, notices, and guidelines) ‘do not offer sufficient guidance’.10 According to Article 15 of Regulation 1/2003, it is entirely within the discretion of national courts to decide whether to ask for this form of assistance. Moreover the requesting court is not formally obliged to conform its decision to the outcome of this assistance.11 The reason for this is the fact that the Commission’s assistance is given without any requirement that the parties are heard by the Commission or are allowed to produce evidence. If the national court was bound by the Commission’s opinion, or would blindly follow it, this would rise serious problems of due process.12 The present practice of cooperation between national courts and the Commission is positive. Within five years (from may 2004 until April 2009), the Commission has provided opinions on eighteen occasions.13 7

Commission Notice on Cooperation between the Commission and Courts of the EU Member States in the Application of Articles 81 and 82 EC, OJ 2004 C101/54, n. 11, par. 21-30. 8 Commission Notice on Cooperation between the Commission and Courts of the EU Member States in the Application of Articles 81 and 82 EC, OJ 2004 C101/54, Par 22. 9 Commission Notice on Cooperation between the Commission and Courts of the EU Member States in the Application of Articles 81 and 82 EC, OJ 2004 C101/54, Par. 28. 10 Commission Notice on Cooperation between the Commission and Courts of the EU Member States in the Application of Articles 81 and 82 EC, OJ 2004 C101/54, Par 27. 11 Commission Notice on Cooperation between the Commission and Courts of the EU Member States in the Application of Articles 81 and 82 EC, OJ 2004 C101/54, Par 19 and 29. 12 Gilliams [2003], p. 462. 13 See Commission Staff Working Paper Accompanying the Communication from the Commission to the European Parliament and Council on the Report of the Functioning of Regulation 1/2003, SEC(2009) 574 final, par. 277-282.

151

4. ARBITRAL TRIBUNALS’ POWERS TO APPLY EU COMPETITION LAW

4.1.2

Arbitral Tribunals and the Cooperation Mechanism of Regulation 1/2003

The legal literature explains that it is not easy to answer the question whether arbitral tribunals have the power to request the assistance from the European Commission or for that matter from any other National Competition Authority.14 Arbitrators are private adjudicators. By definition they have to be neutral and independent from the State intervention as well as from the one of an administrative authority. Unlike national courts, arbitrators are not directly subject to the third paragraph of Article 4 TEU.15 Hence Regulation 1/2003 has not necessary overlooked to mention arbitration. Legal scholars use this argument to affirm that the Cooperation Mechanism of Regulation 1/2003 is not transposable to arbitration.16 Therefore, Article 15(1) of Regulation 1/2003 should not be used as the legal ground to establish whether arbitrators have the power to seek assistance from the Commission or whether the Commission has the duty to reply. Furthermore, arbitral tribunals are also excluded from the Cooperation Notice.17 In fact, instead of referring to court or tribunal as Article 267 TFEU does, it refers only to court and in its first paragraph states that: ‘For the purpose of this notice, the courts of the EU Member States (hereinafter national courts) are those courts and tribunal within a EU Member State that can apply Article 81 and 82 EC [now Article 101 and 102 TFEU] and that are authorized to ask for a preliminary question to the Court of Justice of the European Communities pursuant to Article 267 TFEU.’18 Nevertheless, as Komninos argues, given the soft-law nature of the Coopera14

Blanke and Landolt [2010], Chapter 21 by Assimakis P. Komninos. Komninos [2001]; Article 4(3) reads as follows: ‘Pursuant to the principle of sincere cooperation, the Union and the Member States shall, in full mutual respect, assist each other in carrying out tasks which flow from the Treaties’. 16 Radicati di Brozolo [2004]; Liebscher [2003a]. 17 Commission Notice on Cooperation between the Commission and Courts of the EU Member States in the Application of Articles 81 and 82 EC, OJ 2004 C101/54 18 Commission Notice on Cooperation between the Commission and Courts of the EU Member States in the Application of Articles 81 and 82 EC, OJ 2004 C101/54, par. 1. 15

152

tion Notice, although there will be no power-duty relationship between arbitrators and the Commission, the mechanism could still be used by analogy to help arbitral tribunals apply EU competition law. ‘It is reasonable to believe that the Commission intended to exclude arbitration only from the specific procedural framework of the New Cooperation notice, while entertaining requests from arbitrators on an ad hoc and fully discretionary basis, rather than being bound to engage in a dialogue with arbitrators as it is bound to do so with courts.’19 In fact it will be illogical for the Commission to neglect or ignore its long-established favour for antitrust arbitration.20 Moreover if the Commission was to leave arbitrators unassisted in the application of competition law, they could rather ignore a complex antitrust issue instead of risking that the whole award is set aside during the second-look review. Furthermore, the Commission has a central role in the enforcement of EU competition law and it will be inconsistent with its institutional role to overlook the private antitrust enforcement in arbitration proceedings. Finally, if there is an informal cooperation mechanism with arbitral proceedings seated within the EU, we do not see any reason why this should not be available also to arbitral proceedings seated in third countries. Regarding the definition of the assistance that arbitrators could informally request and receive from the Commission, it could be comparable to the one provided to national courts, namely:21

• factual information, for instance, economic data, such as statistics, market characteristics, and economic analyses; • information on whether a certain case is pending before the Commission; • whether the latter has reached a decision or a reasonable opinion in this matter; and 19

Blanke and Landolt [2010], Chapter 21 by Assimakis P. Komninos. See the Introduction Chapter of this work. 21 Zuberb¨ uhler and Oetiker [2007], P. Peyrot, Expert Determination of Competition Issues, p. 109. 20

153

4. ARBITRAL TRIBUNALS’ POWERS TO APPLY EU COMPETITION LAW

• a legal issue of EU competition law.22

The final issue of interest for this description of the cooperation between arbitrators and the Commission are the requirements that make it appropriate for the arbitrators to seek this assistance. It is worth noting that arbitration is a private proceeding often subject to confidentiality. Hence, the cooperation between arbitrators and the Commission raises serious concerns with respect to confidentiality. Although this topic is out of the scope of the present work, the parties could authorise the arbitral tribunal to use this tool waiving the confidentiality clause. However it may be difficult to reach the parties’ agreement after the dispute started, so it will be wise to include a provision on this matter in the contractual arbitration clause. Nevertheless it may also be difficult for the parties to negotiate such a provision ex ante. A wise alternative would be for arbitration institutions and for jurisdictions to include in their arbitration rules a provision leaving to the arbitrators the freedom to decide sua sponte whether to request the Commission assistance or not.

To conclude, there is no legal provision that can be used as legal basis to claim that arbitral tribunals’ have the formal power to request the assistance of the Commission, forcing the latter to comply. However, it is possible and desirable that the cooperation rules valid for national courts are indirectly applied also to international arbitral tribunals. To reconcile arbitration with this cooperation procedure with an administrative authority it will be wise to have parties’ agreement on this point, out of respect for the fundamental principle of party autonomy. Alternatively legal systems and arbitration institutions could include into their lex arbitri procedural rules a provision on this matter.

22

This last point is consistent with the outcome of the analysis of the second Section of this Chapter, below, and its meaning will be discussed later.

154

4.2

Arbitral Tribunals’ Jurisdiction to Request a Preliminary Ruling Under Article 267 TFEU

The goal of this Section is to analytically test whether and how the case-law of the European Court of Justice on this specific procedural rule affects individual’s ex ante incentives to exclude competition law from a contractual arbitration clause. The CJEU’s interpretation of Article 267 TFEU denies the arbitral tribunals’ jurisdiction to refer cases for a preliminary ruling. Hence it risks to increase arbitration litigation costs in comparison to the national courts litigation costs. Moreover it risks to deprive an arbitral award based on EU competition law from its effectiveness. To verify this statement we adapt the Priest and Klein model23 on the selection of disputes for litigation and integrate it within Dari-Mattiacci and Parisi’s model on dissipation of value and lost treasure in rent-seeking games.24 There are three preliminary remarks we need to address to justify the economic analysis used. The first is the functioning of Article 267 TFEU in light of the CJEU’s interpretation, the second regards the limitations to the recognition and enforcement of an arbitral award by a national court and the third describes why denying arbitral tribunals’ authority to request a preliminary ruling may impair the effectiveness of arbitration when European competition law is involved. Subsequently, the participation constraint is analysed, since it is a fundamental building block for describing how the legal rule stated in CJEU’s case-law on Article 267 TFEU affects plaintiff’s expected utility function when he has to decide whether to file or not to file an arbitration proceeding that potentially involves European public policy issues. This Chapter will be concluded with the parties’ effort analysis, testing how parties’ level of effort in winning a dispute is affected by the CJEU’s case-law on Article 267 TFEU. Through this economic analysis we obtain a theoretical model to find whether the rule created by the CJEU’s case-law can cause an increase in lost treasure in comparison with the opposite interpretation of Article 267 TFEU supported by legal scholars. Such uncertainty and the existence of a potential higher lost treasure in arbitration 23 24

Priest and Klein [1984]. Dari-Mattiacci and Parisi [2005].

155

4. ARBITRAL TRIBUNALS’ POWERS TO APPLY EU COMPETITION LAW

could provide parties with the incentive to avoid arbitration and go back to traditional litigation, although the latter has so far proved to be an unsatisfactory tool for the private enforcement of competition law.

4.2.1

The Functioning of Article 267 TFEU

According to Article 19 of the Treaty on the European Union the function of the Court of Justice of the European Union is to ‘ensure that in the interpretation and application of the Treaties the law is observed’ and in doing so the CJEU shall ‘give preliminary rulings, at the request of courts or tribunals of the Member States, on the interpretation of Union law’.25 The procedure for requesting a preliminary ruling is described by Article 267 TFEU: ‘The Court of Justice of the European Union shall have jurisdiction to give preliminary rulings concerning: (a) the interpretation of the Treaties; (b) the validity and interpretation of acts of the institutions, bodies, offices or agencies of the Union; Where such a question is raised before any court or tribunal of a Member State, that court or tribunal may, if it considers that a decision on the question is necessary to enable it to give judgement, request the Court to give a ruling thereon. Where any such question is raised in a case pending before a court or tribunal of a Member State against whose decisions there is no judicial remedy under national law, that court or tribunal shall bring the matter before the Court.’ According to this provision, through its preliminary ruling, the CJEU is the exclusive judicial authority with the power to provide a legally binding interpretation of the European Union law.26 However, this power can be exercised only when a court or tribunal of a Member State decides that a preliminary ruling is necessary to give its judgement. Moreover, the CJEU has no power to affect this decision nor to act as an appellate court assessing the merit of the dispute.27 25

Article 19 of the Treaty on European Union. More specifically the CJEU ruling is legally binding only for the court that made the request of preliminary ruling. However the whole European legal system takes under deep consideration CJEU’s case-law.Craig and De B´ urca [2011], pp. 442-448. 27 As correctly argued in Hartley [2007], p. 287, ‘There are two major differences between an appeal and a reference. First, in the case of an appeal the initiative lies with the parties: the party who is dissatisfied with the courts judgement decides whether to appeal and then take 26

156

It just provides the correct interpretation of the Union law in the case at stake. Using the words of the CJEU: ‘The objective of the procedure is to retain the independence of the national courts, while at the same time preventing a body of national case law not in accord with the rules of (the European Union) law from coming into existence in any Member State guaranteeing an uniform application of EU law throughout the Union’.28 It is worth noting that the preliminary ruling procedure accounts for over 50% of all cases heard by the CJEU, making Article 267 TFEU a key feature in the development and enforcement of the European Union law. After this description of Article 267 TFEU and how it works towards Member State National Courts, our attention can shift on CJEU’s case-law on this article in relation to antitrust arbitration. More specifically, how this provision works, according to the CJEU, when private arbitral tribunals are handling a dispute that requires a preliminary ruling on articles 101 and 102 of the TFEU. The issues is whether an arbitral tribunal can be considered ‘court or tribunal of a Member State’ in the meaning of Article 267 TFEU and if it is not the case, the question is how arbitral tribunal are supposed to act in such circumstances to solve the dispute. The first decision on this matter is the Nordsea case, dated 1982.29 Here, the CJEU denied private arbitral tribunals jurisdiction to refer for preliminary rulings stating that: ‘An arbitrator who is called upon to decide a dispute between the parties to a contract under a clause inserted in that contract is not to be conthe necessary procedural steps. The court a quo normally has no further say in the matter and cannot prevent the appeal from being lodged. Secondly, the appeal court decides the case, even though the appeal may be on limited grounds only, and it has the power to set aside the decision of the court a quo; normally it can then substitute its own decision for that of the lower court. These features are not found in the procedure for a preliminary reference: the court a quo decides whether the reference should be made, and only specific issues are referred to the European Court. Once it has decided these, the European Court remits the case to the national court for a decision.’ 28 Case 107/76, Hoffmann-La Roche AG v Centrafarm Vertriebsgesellschaft Pharmazeutischer Erzeugnisse mbH [1977] ECR 9576. 29 Case C-102/81, Nordsee Deutsche Hochseefischerei GmbH v Reederei Mond Hochseefischerei Nordstern AG & Co KG [1982] ECR 1095.

157

4. ARBITRAL TRIBUNALS’ POWERS TO APPLY EU COMPETITION LAW

sidered as a ’court or tribunal of a Member State’ within the meaning of Article 177 of the Treaty [now called Article 267 TFEU]’.30 However, the CJEU did not leave a procedural gap for such circumstances. In fact it also explained how the preliminary ruling shall be requested: ‘If in the course of arbitration resorted to by agreement between the parties questions of community law are raised which the ordinary courts may be called upon to examine either in the context of their collaboration with arbitration tribunals or in the course of a review of an arbitration award, it is for those courts to ascertain whether it is necessary for them to make a reference to the Court of Justice under Article 177 of the Treaty [now called Article 267 TFEU]’.31 Substantially, this decision entitles national courts to have a second-look on arbitral tribunals’ award and to examine their application of European Union Law in the context of any method of recourse available under the relevant national legislation. It also totally deprives arbitral tribunal of jurisdiction on the necessity to refer to the Court of Justice for a preliminary ruling. Several legal scholars started a debate on the CJEU’s interpretation of Article 267 TFUE and they showed how, from a legal perspective, there is room to strongly argue both pro and against arbitrator’s jurisdiction on preliminary ruling. From this debate it is clear that the issue is controversial and the wording of the TFEU allows to reverse the interpretation taken in CJEU’s case-law.32 However, in 1994 the CJEU confirmed its interpretation and build further in the case Municipality of Almelo.33 It stated that: ‘It follows from the principles of the primacy of Community law and of its uniform application, in conjunction with Article 5 of the Treaty, that a court of a Member State to which an appeal against an arbitration award is made pursuant to national law must, even where it gives judgement having regard to fairness, observe the rules of Community law, in particular those relating to competition (law).’34 In this specific case it was 30

Case C-102/81, Nordsee Deutsche Hochseefischerei GmbH v Reederei Mond Hochseefischerei Nordstern AG & Co KG [1982] ECR 1095. 31 Case C-102/81, Nordsee Deutsche Hochseefischerei GmbH v Reederei Mond Hochseefischerei Nordstern AG & Co KG [1982] ECR 1095. 32 Biavati [1995]; Tizzano [1983]. 33 Case C-393/92 Gemeente Almelo et al. v. NV Energiebedrijf IJsselmij [1994] ECR I-1277. 34 Case C-393/92 Gemeente Almelo et al. v. NV Energiebedrijf IJsselmij [1994] ECR I-1277, par. 23.

158

possible to appeal the award but the national court had to decide the case not using the body of national legislation but having regard to the principle of fairness. The fundamental legal principle implied in Municipality of Almelo is that no matter what is the substantial law freely chosen by the parties to regulate the merit of the dispute, the national court has to guarantee that the EU law is observed and in particular that European competition law is observed. This is a ground-breaking principle for the review of arbitral awards by national courts. And it was clarified five years later in the Eco Swiss case.35 In Eco Swiss the CJEU starts quoting its Nordsea decision but, in light of scholars’ criticisms that followed the too wide principle expressed in Municipality of Almelo, it proceeds specifying how ‘it is in the interest of efficient arbitration proceedings that review of arbitration awards should be limited in scope and that annulment of or refusal to recognise an award should be possible only in exceptional circumstances’36 but then, the CJEU states that: ‘according to Article 3(g) of the EC Treaty (now, after amendment, Article 3(1)(g) EC), Article 85 of the Treaty (now called Article 101 TFEU, a competition law provision) constitutes a fundamental provision which is essential for the accomplishment of the tasks entrusted to the Community and, in particular, for the functioning of the internal market.’37 For this reason, according to the CJEU, a national court must set aside an arbitral award if the court considers that the award is in fact contrary to Article 101 TFEU, where its domestic procedural rules allow the court to set it aside on the ground of failure to observe national rules of public policy.38 The CJEU also states that: ‘the provisions of Article 85 (now called Article 101 TFEU) of the Treaty may be regarded as a matter of public policy within the meaning of the New York Convention.’39 To clarify the extent and fully understand the 35 36

Case C-126/97, Eco Swiss China Time Ltd v Benetton International NV [1999] I-3055. Case C-126/97, Eco Swiss China Time Ltd v Benetton International NV [1999] I-3055, par.

35. 37

Case C-126/97, Eco Swiss China Time Ltd v Benetton International NV [1999] I-3055, par.

36. 38

Case C-126/97, Eco Swiss China Time Ltd v Benetton International NV [1999] I-3055, par.

37. 39

Case C-126/97, Eco Swiss China Time Ltd v Benetton International NV [1999] I-3055, par. 38; The New York Convention of 10 June 1958 on the Recognition and Enforcement of Foreign Arbitral Awards.

159

4. ARBITRAL TRIBUNALS’ POWERS TO APPLY EU COMPETITION LAW

importance of this statement we have to shift our attention on the limitations to validity of arbitral award.

4.2.2

Limitations to the Recognition and Enforcement of an Arbitral Award

All the European Union Member States are among the 142 United Nations Member States, that have signed and adopted the New York Convention of 10 June 1958 on the Recognition and Enforcement of Foreign Arbitral Awards, making it one of the most widely adopted international law conventions.40 The object of the New York Convention is to guarantee that an award granted by a foreign arbitral tribunal can be recognised and enforced in all the countries that have adopted the Convention. There is no need to explain why a widely enforceable arbitral award is a fundamental aspect for the functioning and efficiency of arbitration. In fact, as already mentioned, the CJEU in Eco Swiss acknowledges the general principle of limited review of arbitral awards: ‘it is in the interest of efficient arbitration proceedings that review of arbitration awards should be limited in scope and that annulment of or refusal to recognise an award should be possible only in exceptional circumstances’.41 In Article V(1)(c) and (e) and II(b) of the New York Convention we can find the limitation to the general principle of recognition and enforcement of arbitral awards. More specifically, according to letter b), any national court has the power to set an arbitral award aside, under the condition that the recognition or enforcement of the arbitral award would be contrary to the public policy of that country. The same limitation is often present within national procedural regulations and it reflects the French general principle of “ordre public”.42 To some extent this principle allows a form of action or review of a national arbitral award with the purpose to control how national mandatory norms are applied in arbitration and 40

As of October 1, 2009. Biavati [1995]; Tizzano [1983]. 42 See for instance: Article 1064 and 1064 of the Dutch Code of Civil Procedure; Article 829 and Article 839 of the Italian Code of Civil Procedure; Article 1502 of the New French Code of Civil Procedure; Article 187 and 190, paragraph 2, lett. e) of the Swiss “Loi fdrale sur le droit International priv”. 41

160

most importantly to impede the enforcement of a decision which is contrary the national public policy.43 However, the legal literature on arbitration acknowledges that the purpose of this limitation is not to provide an appeal judgement in front of national courts.44 It represents the last resort mechanism to avoid enforcing an award that results incompatible with the fundamental principles of a national legal system and that can be considered against the the national public policy provisions.45

4.2.3

The Importance of Preliminary Ruling for Arbitration, When “Public Policy” is at Stake

In light of what said in 4.2.1 and 4.2.2 above, we can now understand the importance of the aforementioned Eco Swiss case.46 The CJEU substantially declared Article 101 TFEU a matter of public policy, with respect both to the review of national arbitration and to recognition-enforcement of foreign arbitral awards under the New York Convention. Therefore, provided that an arbitral award involves European competition law, arbitrators have no jurisdiction over issues of preliminary ruling and a national court has to inquire whether a preliminary ruling is necessary to decide the case. If the national court deems the preliminary ruling necessary, it can refer the case to the CJEU under Article 267 TFEU. After the CJEU gives its preliminary ruling, the national court sets the arbitral award aside, unless the arbitral tribunal was capable ex ante, so before the preliminary ruling, to interpret and apply the EU law at stake exactly as interpreted afterwards by the CJEU in its preliminary ruling. When an arbitral tribunal is denied jurisdiction on reference for preliminary ruling and the same tribunal is facing a legal dispute that involves the private enforcement of European competition law, the effectiveness of the arbitral award is on the edge. Arbitrators have no way to prevent a national court from setting 43

The concepts of mandatory norms and public policy rules are two different concepts but in the legal practise sometimes for judges is difficult to set the boundary. 44 Derains [1987]. 45 Derains [1987] 46 Case C-126/97, Eco Swiss China Time Ltd v Benetton International NV [1999] I-3055.

161

4. ARBITRAL TRIBUNALS’ POWERS TO APPLY EU COMPETITION LAW

aside their decision. The paradoxical result is that arbitrators have the power to decide whether mandatory norms are applicable or not and at the same time the arbitral tribunal may not assess whether a preliminary ruling is needed. It has no jurisdiction on the point, and no matter how the tribunal applies the EU law any national court may deem that a preliminary ruling was indeed necessary and refer the case to the CJEU. It is worth repeating that the procedure in front of a national court cannot represent an appeal procedure. Namely, an assessment on how arbitrators applied the law to the merit of the case is out of the scope of the national court’s second-look procedure. On the contrary, national courts have just to guarantee that European competition law was duly taken into consideration in the arbitration proceeding and if it is not the case to set aside the award.

On the ground of these three preliminary remarks we can now move to the economic analysis, but first we should mention that to render this problem of a more general interest, it is worth noting that, beyond competition law, it is uncertain which other provisions of the EU Treaties are eligible to be considered European “public policy”. At any time, the CJEU may declare a EU law provision to be of fundamental importance for the EU legal order casting a shadow on all arbitration proceedings involving that provision.

4.2.4

Participation Constraint Analysis

In order to be able to compare the two possible interpretations of Article 267 TFEU, in a Law and Economics efficiency seeking perspective, we need a more structured and simplified definition of the problem and an outline of the general assumptions necessary for the economic analysis. Then, once the problem definition is complete and given some useful general assumptions, our attention will shift ot private parties’ ex ante expected utility in national court litigation which is described in the Priest and Klein model. This model is a cornerstone of litigation theory and is suitable for studying the preliminary reference procedure applicable to national courts.

162

4.2.4.1

Description of the Structure of the Problem

In the EU legal order the lawmaker can choose between two procedural legal rules to apply when an arbitral tribunal considers a preliminary ruling from the CJEU necessary to be able to correctly interpret and apply EU competition law to grant an enforceable award. Our intention is to test and compare the two legal solutions that are theoretically available under Article 267 TFEU. One is the CJEU’s current interpretation of Article 267 TFEU and the other is the alternative interpretation supported by legal scholars which grant arbitrators the same powers that national courts have, with respect to a request of preliminary ruling. We call them “legal rules” but they are actually the two possible interpretations that can be given of Article 267 TFEU with respect to private arbitration proceedings. In order to analyse how the two legal rules differ and whether one is better than the other, it is necessary to provide a more structured and simplified description of them:

1. The rule granting arbitral tribunals jurisdiction to request a preliminary ruling is exactly identical, from a procedural viewpoint, to the one applicable to private enforcement in front of national courts.47 In this case arbitrators will suspend the procedure and request the preliminary ruling to the CJEU, as national judges do. After receiving the CJEU’s decision on the preliminary ruling an arbitral tribunal can grant an award in compliance with the application of the European “public policy” rules. Such arbitral award will be difficult to overrule by national courts in the enforcement phase on the grounds of European “public policy” issues. In fact, national courts’ review will concern only whether the issues on application of mandatory norms and preliminary ruling were duly taken into consideration; 2. Alternatively, as per the state of the art of the CJEU’s interpretation of article 267 TFEU, if the arbitral tribunal is denied authority to request preliminary rulings, the award is on the edge. Arbitrators have two options: 47

For the sake of simplicity, when we will be referring to this rule we will interchangeably use the expression the national courts procedure or arbitrators’ jurisdiction/power to request a preliminary ruling.

163

4. ARBITRAL TRIBUNALS’ POWERS TO APPLY EU COMPETITION LAW

(a) they can suspend the arbitration procedure requesting the national court to assess the preliminary ruling issue, under a cooperation scheme with a national courts when it is available under the national procedure, or (b) they can grant an arbitral award irrespective of the need of a preliminary ruling to apply European “public policy”. The latter case implies that, in light of the CJEU case-law on article 267 TFEU, the award can be voided by any Member State national court on the ground either of national procedural law or of the Article V paragraph 2, let b) of the New York Convention. Eventually the plaintiff will have to file a second arbitral procedure to obtain an enforceable award. In what follows, our goal is to test each rule with respect to its effect on parties’ ex ante incentives to choose to exclude competition law disputes from a contractual arbitration clause. 4.2.4.2

General Assumptions

A first assumption is based on the fact that a CJEU preliminary ruling does not concern the merit of the case but it is only a matter of EU law interpretation. Hence, we can assume that parties’ effort in litigation cannot directly influence the outcome of the preliminary ruling.48 However, parties’ investment in litigation effort can influence both the decision of the arbitral tribunal on the merit of the case and the decision of the national court on whether a preliminary ruling is necessary. We can also assume that under the second rule (the lack of authority to request a preliminary ruling) there is no overlap of respective jurisdiction between arbitral tribunals and national courts. In fact, as mentioned above, on the one hand it is a general principle of arbitration that national courts cannot review an arbitral award on the merit of the case. On the other hand, also arbitrators, even when cooperating with a national court, can express only opinions on the matter of preliminary ruling and do not have any jurisdiction to address this issue. 48

Dari-Mattiacci and Parisi [2005].

164

4.2.4.3

Definition of the participation constraint

We can focus now on the economic analysis of the two legal rules discussed above. The first rule, which we call Ra , is applicable to national courts and if made available to arbitrators will provide arbitral tribunals with the authority to request preliminary rulings. Under the Ra rule, parties’ ex ante participation constraint to the arbitration proceeding is described by the Priest and Klein model:

Ra : pa W − (1 − pa )L − C − S > 0

(4.1)

With pa representing plaintiff’s estimation of the probability of victory; W is the amount at stake in case plaintiff wins; (1 − pa ) is plaintiff’s estimation of the probability of defendant’s victory; L is plaintiff’s loss in case defendant wins a possible counterclaim. The term C represents the cost of the procedure and S is the opportunity cost of settlement. Only when this participation constraint is positive, so Ra > 0, plaintiff will file a lawsuit.49 This well known process of selection of disputes for litigation needs to be modified if we focus our attention on what happens under the second rule, which we call Rc , namely when arbitral tribunals are denied authority to request a preliminary ruling. As clarified above, under the rule Rc , the parties are facing a three-phase procedure and according to our assumptions the adjudicator of each phase has jurisdiction only over legal issues that are specific to its own phase. In other words, the three adjudicators are three different bodies with independent jurisdictions. Therefore, under the Rc rule, the participation constraint is not equal to Ra any more. Plaintiff will be facing a differently structured decision 49

Priest and Klein [1984] The standard participation constraint indicated in the literature is: pW − C > S and means that plaintiff’s expected gain from litigation is bigger than the potential settlement amount. However we are not interested in the settlement, so we prefer a formulation that considers settlement as a opportunity cost of litigation. Moreover, the formulation we choose here considers also the fact that defendant may have counterclaims, for instance reputation damages.

165

4. ARBITRAL TRIBUNALS’ POWERS TO APPLY EU COMPETITION LAW

problem. In fact plaintiff’s probability to win the case is represented by three different and independent probabilities relative to the first, the second or the third phase of the procedure, which we call respectively p1 , p2 and p3 . Given the fact that these three probabilities are attached to three different and completely independent events (see what stated in Section 4.2.3), to find plaintiff’s total probability of victory we multiply the independent probabilities of each independent event. We repeat, for the sake of clarity, that the third phase is a re-file of the arbitration, which is needed if the first arbitral award is set aside by a national court in the second phase. Figure 4.1, below represents all the possible outcomes of the multi-phase dispute with the relative probabilities and identifies those positive for plaintiff as well as those positive for defendant.

Figure 4.1: Probabilities attached to all possible outcomes of a CJEU preliminary ruling request in arbitration

166

Figure 4.1 is telling us that during Phase 1, plaintiff has probability p1 to get a positive outcome in the arbitration phase with respect to the merit of the dispute, and probability 1 − p1 to get a negative outcome. In both cases, during Phase 2 plaintiff has probability p2 to get a positive outcome in front of the national court on the procedural issue about whether a preliminary ruling ex Article 267 TFEU to the CJEU is necessary to solve the dispute as well as whether after the referral the arbitral award needs to be set aside.50 Finally p3 is the probability that eventually plaintiff wins Phase 3, namely the new arbitration procedure which comes after the national court’s referral for preliminary ruling, the CJEU response and the set aside of the first arbitral award by the national court. Each ending position represents a possible outcome of the multi-phase dispute that may be favourable or unfavourable to plaintiff. The expression Rc , below, 50

The reason why the same probability p2 is attached to plaintiff’s request for a preliminary ruling on one side of the Figure 4.1 and on the other side of the same Figure it is attached to plaintiff’s opposition of a preliminary ruling, may not be of immediate understanding. The reason for this lies on the fact that neither the merits of the case nor plaintiff’s interpretation of the applicable legal rule do not change according to the outcome of Phase 1. This is based on the fact that as explained above, the three phases are linked to independent events. In Phase 1 the parties are disputing the specific merits of the case. In order to decide the case the arbitral tribunal needs to interpret the EU legal rules at stake. In Phase 2 the parties are disputing over the correct interpretation of the rules at stake. Therefore, these two independent events are based on different and independent matters. More specifically, the probability p2 is unique because it is associated to a unique legal argument made by plaintiff regarding the correctness of his interpretation of the law. Depending on the outcome of Phase 1, this unique legal argument is simply attached to two different demands in the two branches of Figure 4.1. In fact, on the one hand, when plaintiff wins Phase 1, in Phase 2 he is going to argue that his interpretation is the correct interpretation of the legal rule at stake, hence there is no need of a preliminary ruling. On the other hand, when plaintiff loses in Phase 1, he is still going to argue that his interpretation is the correct one in Phase 2. However in this last case, plaintiff needs to demand to the national court to refer the case for a preliminary ruling because of the state of the proceedings. In any case, these are just different concrete requests to the judge that plaintiff has to make to support his unique legal argument in the dispute. The soundness of his legal interpretation and the attached probability does not change in relation to the outcome of Phase 1. This Section is based on the fact that arbitrators cannot determine whether plaintiff’s interpretation of the law is correct or whether it is uncertain, hence needing a preliminary ruling. No matter which interpretation the arbitral tribunal applies in Phase 1, the national court has to examine which is the correct interpretation of the legal rule at stake. When the national court’s interpretation corresponds to the arbitrators’ one, there will be no preliminary ruling, otherwise there will be a request for preliminary ruling. Nevertheless the national court does not decide whether the arbitral tribunal interpretation is correct or not. It has to decide this matter autonomously and independently.

167

4. ARBITRAL TRIBUNALS’ POWERS TO APPLY EU COMPETITION LAW

represents how the participation constraint of Equation 4.1 needs to be redefined to describe what happens in this scenario:

Rc : [p1 p2 + p1 (1 − p2 )p3 + (1 − p1 )p2 p3 ]W −[(1 − p1 )(1 − p2 ) + (1 − p1 )p2 (1 − p3 ) + p1 (1 − p2 )(1 − p3 )]L − C − S > 0

After calculating and simplifying the participation constraint, the following result is reached:

Rc : (p1 p2 + p2 p3 + p1 p3 − 2p1 p2 p3 )W −(1 − p1 p2 − p2 p3 − p1 p3 + 2p1 p2 p3 )L − C − S > 0

(4.2)

This multi-phase procedure affects the participation constraint in two ways: on the one hand, the outcome depends on different and independent events that must occur simultaneously, in fact we multiply the probabilities affecting each outcome. On the other hand, in this scenario each party has the opportunity to revert an unfavourable decision, in fact to find the total parties’ probability of success we have to sum the probabilities of each outcome. Therefore, in our model, the two legal rules under investigation are represented respectively by Equation 4.1 and 4.2, above. To show the consistency of our model with the Priest and Klein model, in Equation 4.2 we can set: p1 p2 + p1 p3 + p2 p3 − 2p1 p2 p3 = pc the two expressions will have the same structure:

Ra : pa W − (1 − pa )L − C − S > 0 Rc : pc W − (1 − pc )L − C − S > 0

) ∀pc = p1 p2 + p1 p3 + p2 p3 − 2p1 p2 p3

168

This statement may appear trivial but it confirms the suitability of our problem settings and it is a building block for the economic analysis of the two legal rules and for further research on this topic. A possible critique may involve the introduction of a the third phase, since it is substantially equal to a redo of the first phase in light both of the national court decision and of the CJEU’s one. We believe that such critique is not sound. The third phase is actually connected only to the CJEU’s decision which is independent from the previous two phases and from the merit of the case. If the third phase is reached, this means that the CJEU’s preliminary ruling was in contrast with the first phase arbitral award, resulting in the national court setting the arbitral award aside. The third phase is like a new dispute between the same parties on the same issue but in a new state of the world changed by the CJEU’s preliminary ruling. In fact, in the third phase, the merit of the dispute must still be decided by an arbitral tribunal (not necessary the same tribunal of the first phase) legally obliged to comply with the outcome of the CJEU’s interpretation of European competition law rule. The reasoning above holds true also when a national regulation setting a cooperation procedure between arbitrators and national courts exists, so when the arbitral tribunal can suspend the first phase procedure and request a national court to deal with the preliminary ruling. Later on, when the arbitral tribunal resumes the first stage, it will address the merit of the case exactly as mentioned above. This scenario will substantially be a three-phase procedure compatible with the one described above. At this point, further research should be driven by considerations about individuals perception and estimation of the uncertain probability of winning in a multi-phase procedure. In fact Priest and Klein showed that ‘disputes selected for litigation will constitute neither a random nor a representative sample of the set of all disputes’.51 They will be those disputes presenting two characteristics: 51

Priest and Klein [1984]

169

4. ARBITRAL TRIBUNALS’ POWERS TO APPLY EU COMPETITION LAW

(a) a probability of success estimated by the parties which is closer to the threshold probability and (b) the difference between the parties’ expectations of the outcome is bigger. Hence, only those cases where parties are more likely to make an error in the assessment of the probability of success will end up in litigation.52

4.2.5

Dissipation of Parties’ Effort Analysis

Focusing now on parties’ efforts in each phase of the procedure, the goal is to discover whether private parties’ can ex ante expect that different levels of rentseeking efforts will be needed in a multi-phase arbitration procedure in comparison to national courts litigation proceedings. If this is the case, then one legal rule is causing a higher lost treasure in litigation. Therefore private parties behind the veil of ignorance, having to decide ex ante between arbitration an ordinary litigation, will be subject to incentives to choose the private enforcement mechanism that has lower transaction costs. In order to study the transaction cost impact of the two legal rules discussed above we need a Law and Economics tool. In this case it is the standard formulation of a rent-seeking game in litigation:

αAr p= αAr + βB r

According to the theory of rent-seeking in litigation, the probability of winning the case p, is a function of A, that is plaintiff’s level of effort in litigation, and B, defendant’s level of effort in litigation; with α and β representing coefficients of parties respective position in the merit of the dispute. The term r is a factor determining the productivity of rent-seeking expenditures. For the sake of simplicity, we shall assume that parties’ stakes in the dispute are symmetric (W = L), in fact this is often the truth in contractual relationships. The model we are using to solve this problem is an adaptation of the one used by Dari52

Priest and Klein [1984]

170

Mattiacci and Parisi to describe the rise of rent dissipation. We also choose to follow part of their notation and the useful simplification through normalisation of the value at stake W .53 In Dari-Mattiacci and Parisi’s model each party’s share of the rent seeking competition (RA for plaintiff and RB for defendant) will be equal to:

RA = pA − A RB = pB − B

With pA and pB representing parties’ respective probabilities to win the case in function of their respective efforts. As mentioned above, the parties value at stake W is not present in the formula because of the normalisation. Assuming for simplicity that the parties have also symmetric merits in the dispute (namely α = β = 1), the dissipation of effort analysis described by Dari-Mattiacci and Parisi in their model of rent-seeking dissipation is well suited to be applied when the first rule Ra is in force. According to this model, under rule Ra described above, private parties can expect that in litigation their net share of the rent is going to be:54 53

Dari-Mattiacci and Parisi [2005], ‘In order to keep notation simple, we model the parties’ investment in rent-seeking as fractions of the rent. The literature has usually used capital letters like X and Y to denote the absolute values of the parties’ expenditures in rent-seeking and X/(X + Y )V and Y /(X + Y )V to denote the parties’ shares in the rent. Party Xs payoff thus is generally represented as X/(X + Y )V − X (the portion of the rent he or she earns minus the rent-seeking expenditure). In our model we use the labels A and B to denote the fractions of the rent that are spent on rent-seeking, rather than the absolute values of the expenditures. In our model, therefore A = X/V . This merely represents a different way to measure effort, and it does not impinge upon the generality or the validity of the model. In our model, party As payoff is A/(A + B)V − AV = (A/(A + B) − A)V . The absolute value of the rent, V (in our model this term V is called W ), can thus be simplified (or simply normalised to 1), allowing us to concentrate on the parties expenditures and the magnitude of the dissipation as fractions of the rent. 54 Dari-Mattiacci and Parisi [2005], We use the notation Ra A and Ra B because low-case a indicates that we are under the rule Ra and the upper-case A indicates that the equation refers to plaintiff’s viewpoint while upper-case B regards to defendant’s viewpoint.

171

4. ARBITRAL TRIBUNALS’ POWERS TO APPLY EU COMPETITION LAW

Ar RaA = pA − A = r −A A + Br Br −B RaB = pB − B = r A + Br

And parties will maximise their own share of the rent by choosing A∗ and B ∗ that satisfy the following first order conditions:

δRaA rAr−1 B r = r −1=0 δA (A + B r )2 δRaB rAr B r−1 = r −1=0 δB (A + B r )2

(4.3)

As proved by Dari-Mattiacci and Parisi, the levels of investment that satisfy Equation 4.3 are A∗ = B ∗ = r/4. This means that under the rule Ra , when r = 2 each party will spend exactly 1/2 of the value of the dispute in rent-seeking and the two parties together will dissipate the whole value of the dispute. This is not just theory, empirical studies confirm this theory. In fact the average amount of rent-seeking expenditure in national courts litigation is close to 1/4 of the value of the dispute for each party, with a total lost treasure in litigation equal to almost half of the value of the dispute.55 In order to describe parties’ effort under the alternative rule Rc , which denies arbitral tribunals’ authority to request a preliminary ruling, what is needed is an adaptation of Dari-Mattiacci and Parisi’s model. In fact under Rc , the values of pA and pB are a function of p1 , p2 and p3 which are the probabilities that a party will win a respective phase of the three-phase procedure. Thus, parties respective 55

Data mentioned by Prof. Parisi in an interview, secondary source and reference still needed.

172

probability of winning the case is pA (p1A , p2A , p3A ), and pB (p1B , p2B , p3B ). Going a step forward also p1 , p2 and p3 respectively are functions of parties’ effort in each phase, namely p1 (A1 , B1 ), p2 (A2 , B2 ) and p3 (A3 , B3 ). What follows is the extended mathematical formulation of what is said above:

α2 Ar2 α3 Ar3 α1 Ar1 ; p = ; p = ; 2A 3A α1 Ar1 + β1 B1r α2 Ar2 + β2 B2r α3 Ar3 + β3 B3r β2 B2r β3 B3r β1 B1r ; p = ; p = = 2B 3B α1 Ar1 + β1 B1r α2 Ar2 + β2 B2r α3 Ar3 + β3 B3r

p1A = p1B

In the light of this new definition of pA and pB , it is possible to change accordingly the formulation of Dari-Mattiacci and Parisi’s model and make it compatible with the three-phase procedure defined by the rule Rc , under which arbitrators have no power to request preliminary rulings. It is now possible to use this new definitions of pA and pB in rule Rc to obtain parties net shares of rent under the same rule Rc . This can be done simply substituting this new definition of pA and pB in Equation 4.2. The result is the following:56

RcA : (p1A p2A + p1A p3A + p2A p3A − 2p1A p2A p3A ) − (A1 + A2 + A3 ) RcB : (p1B p2B + p1B p3B + p2B p3B − 2p1B p2B p3B ) − (B1 + B2 + B3 )

(4.4)

To make things less complicated, without losing in precision, we can assume that since parties are not ex ante aware of what will be the merit of the disputes they may assume to have symmetric cases with respect to the merits in all the phases of the procedure, namely that in all p1, p2 and p3 the values 56

Dari-Mattiacci and Parisi [2005], we use the notation Rc A and Rc B because low-case c indicates that we are under the rule Rc and the upper-case A indicates that the equation refers to plaintiff’s viewpoint while upper-case B regards to defendant’s viewpoint.

173

4. ARBITRAL TRIBUNALS’ POWERS TO APPLY EU COMPETITION LAW

α = β = 1. Moreover, under such circumstance, the strategy for a party who wants to maximise its chance to win the game is to divide in three equal amounts the total amount that she is willing to invest in the litigation, namely total effort Atot = A1 + A2 + A3 and Ac = A1 = A2 = A3 , the same will be true for Btot = B1 + B2 + B3 and with Bc = B1 = B2 = B3 .57 Accepting the above assumptions necessarily results in producing equal parties’ probabilities of success in each phase of the dispute, namely p1 = p2 = p3 . These assumptions are not too restrictive. In fact we can expect that in light of the predictions of Priest and Klein model on the selection of disputes for litigation, ex post ‘when either plaintiff or defendant has a “powerful” case, settlement is more likely because the parties are less likely to disagree about the outcome’ of the trial. The contrary is true when a case is problematic, namely when the merit or the law applicable is uncertain.58 Private parties cannot ex ante foresee in which state of the world they will be when in a future moment in time the dispute will arises. Therefore, ex ante parties expect that ex post they will settle cases in which one party has better merits and select for litigation those cases that will actually be posing both the most challenging merits and the most challenging interpretation of competition law. In other words ex ante parties will expect to litigate cases where the probability of winning is close to 50% in all the phases of the procedure and settle otherwise.59 After the simplification introduced by the above assumptions we can rewrite Equation 4.4 defining parties’ net share of the rent in our model as equal to:

RcA = 3p2A − 2p3A − 3Ac ; ∀pA = RcB = 3p2B − 2p3B − 3Bc ; ∀pB =

Arc ; Arc +Bcr Bcr ; Arc +Bcr

57

3r 3A2r c c − (Ar2A r 3 − 3Ac (Arc +Bcr )2 +B c c) 3Bc2r 2Bc3r RcB = (Ar +B r )2 − (Ar +B r )3 − 3Bc c c c c

→ RcA = →

It can be easily proven that from a game theoretical point of view this is the best strategy from an ex ante perspective. 58 Priest and Klein [1984] 59 In further writings, it will be interesting to examine, this issue from the CJEU’s viewpoint which has an institutional mission to make the EU Law application clearer through it’s case-law and therefore has an interest to hear such highly problematic chases.

174

In order to maximise their net share of the rent, for plaintiff with respect to A and for defendant with respect to B, parties will choose A∗c and Bc∗ that satisfy the following first order conditions:60

Bc2r 2rA2r−1 δRcA c = −1=0 δAc (Arc + Bcr )4 2rBc2r−1 A2r δRcB c = −1=0 δBc (Arc + Bcr )4

(4.5)

This result fulfils the requirements set by the Dari-Mattiacci and Parisi’s model.61 In fact, when we compare Equation 4.5 with Equation 4.3, it is clear that they are qualitatively the same functions and they are diverging only quantitatively.62 In fact, it is straight-forward that in Rc , namely under the rule ∗ the total denying authority to arbitrators, when A∗c = Bc∗ , calling A∗tot and Btot levels of parties’ effort in litigation that satisfy Equation 4.5 maximising parties’ ∗ net share of the rent are A∗tot = Btot = r3/8,63 while on the contrary in Ra above, the solution of Equation 4.3, was: A∗ = B ∗ = r/4. In order to clarify the meaning of this statement, it should be reminded that under the rule Ra , when arbitral tribunals have the power to request preliminary 60

See Appendix 1 for the extensive mathematical calculation. Following the notation adopted in this model, obviously we use A∗c and Bc∗ with a low-case c to differentiate this values under rule Rc from A∗ and B ∗ used under rule Ra 61 Second order conditions are always negative for any value of r given that Ac = Bc . 62 In order to test this formulation we can set r = 1 and maximise each share function first with respect to A and then to B obtaining that the model provides results for A∗c = Bc∗ and the values of A and B are positively correlated to r, namely when r increases A and B are going to increase as well. 63 See Appendix 1 for the mathematical calculation of this result. It is based on the description ∗ of the structure of the problem. To obtain the total party’s effort A∗tot and Btot the values A∗c ∗ ∗ ∗ and Bc have to be multiplied by three, because Ac and Bc are the amounts that the parties spend in each phase of the three-phase procedure.

175

4. ARBITRAL TRIBUNALS’ POWERS TO APPLY EU COMPETITION LAW

rulings, each party will spend exactly half of the value of the dispute in rentseeking and the two parties together will dissipate the whole value of the dispute when r = 2.64 On the contrary when the tribunal is denied such authority, so under the rule Rc , the total dissipation is reached at a lower condition r = 4/3.65 The figure below clarifies our results comparing the two rules from a different perspective and representing the total lost treasure that occurs in the dispute resolution under the two analysed rules in function of the variable r:

Figure 4.2: Rent-Seeking Comparison of Alternative Interpretations of Article 267 TFEU

64

See the comment to Equation 4.3 ∗ = 3Bc∗ = 3r/8, we derive that A∗tot = Since, as we mentioned above, A∗tot = 3A∗c and Btot ∗ Btot = 1/2 if and only if r = 4/3. 65

176

After a comment on the graph above we are going to summarise what this model does what all the equations tell us, what the role of the assumptions is and how this model helps to solve the initial problem. To comment the graph above using an example, under the rule granting arbitrators authority to use Article 267 TFEU (called Ra throughout this Section) for r = 1 each party will spend 1/4 of the value of the dispute in litigation with a total lost treasure equal to 1/2 of the total value of the dispute. Under the rule denying arbitrators authority to use Article 267 TFEU (called Rc throughout this Section), for r = 1 each party will dissipate 3/8 of the value of the case,66 with a total dissipation, for the two parties together, equal to 3/4 (simplification of 6/8) of the value of the dispute. Therefore, for r = 1 CJEU’s denied authority on arbitral tribunals will cost parties and society a 25% higher rentseeking dissipation value than what results when arbitral tribunals are granted such authority. The graph shows also that, under the two different rules, full dissipation of the value of the dispute will occur at different values of r. DariMattiacci and Parisi found that when such point is reached if parties have an exit-option, so the option of not filing the lawsuit, parties will start to use the exit option together with mixed participation strategies. ‘When parties randomize their participation in the game, there is a positive probability that no party plays and the rent remains unexploited’ creating a lost treasure loss.67 As we can see from the graph that point is reached sooner and at a lower condition under Rc than under Ra , resulting in less cases of enforcement and more use of exit options. Both our model and Dari-Mattiacci and Parisi’s model intend to describe how much effort, in proportion to the value of the dispute, parties will put into litigation, i.e. an unproductive rent-seeking competition, to maximise their share of the rent, independently of the cost of the procedure itself. Our model differs however, since instead of studying an ordinary litigation procedure it describes what will happen in a three-phase litigation procedure characterised by each phase be66

Namely three times 1/8 of the dispute value which is dissipated in each of the three-phases of the procedure. 67 Dari-Mattiacci and Parisi [2005].

177

4. ARBITRAL TRIBUNALS’ POWERS TO APPLY EU COMPETITION LAW

ing independent from the others. In order to analyse this different problem we had to include a few more key assumptions with respect to Dari-Mattiacci and Parisi. We assume that the three phases of the procedure are independent and we assume that parties will put equal effort and will have equal probability to win each phase; more specifically, in each phase the probability is equal to 50%. The former assumption is derived from the problem definition and the legal analysis; it is the result of a simplification that is necessary to study a legal problem using economic tools. This simplification, however, is not so restrictive with respect to our aims. It is worth reminding that our aim is to qualitatively compare, from parties’ perspective, which of the alternative procedural rules create a more costly enforcement procedure. The latter assumption is derived from Priest and Klein’s assumption that parties will litigate only the disputes with respective merits close to 50%, otherwise parties will settle. Our problem is analysed from an ex ante perspective when parties do not know the merits of each phase and it is consistent with Priest and Klein model. In fact, we are saying that private parties ex ante can assume that, in future disputes, they will litigate only those cases that have the merits of each phase close to 50%. In such scenario the only variable capable to affect the probability of winning is a function of each party’s effort in litigation. Expressing this results in general terms: Proposition 1: Everything else constant, under a rule denying arbitral tribunals’ authority to refer for preliminary ruling, parties expenditure in rent-seeking and the total lost treasure in competition law disputes will be higher in comparison to those under a rule granting arbitral tribunals authority to apply Article 267 TFEU. When parties become aware of this principle, they may ex ante exclude disputes based on competition law from their contractual arbitration clauses preferring the traditional private enforcement litigation forum. The policy implication is quite clear, this is a theoretical argument for reversing the CJEU’s case-law and interpretation of Article 267 TFEU on the basis of an economics analysis of procedural law.

178

To summarise, we applied two of the most famous Law and Economics’ models on litigation to analyse the effects of the European Court of Justice’s interpretation of Article 267 of the Treaty on the Functioning of the European Union. When the European Court of Justice denies arbitral tribunals authority to request a preliminary ruling it also causes adverse effects on parties’ incentives to use arbitration proceedings for resolving their competition law disputes. Consequently it impairs the effectiveness and efficiency of arbitration as a mechanism of private enforcement of EU competition law. The Priest and Klein model on the selection of disputes for litigation helped us to define the participation constraint for entering into litigation under the described CJEU’s case law. However the core of our analysis calls up the Dari-Mattiacci and Parisi model on the dissipation of value through rent-seeking to show how parties’ level of effort is affected by the possible interpretations of Article 267 TFEU. As expected, the result of this exercise provides a strong argument for reverting the current interpretation of Article 267 TFEU. In fact, holding every other variable constant, under the current CJEU’s interpretation of this rule, parties’ dissipation of value through rent-seeking will be higher. In conclusion the investigated CJEU’s case-law needs an amendment if we want to exploit the full potential that arbitration can offer to the private enforcement of EU competition law.

179

4. ARBITRAL TRIBUNALS’ POWERS TO APPLY EU COMPETITION LAW

180

Chapter 5 Arbitral Tribunal’s Duties to Apply EU Competition Law The previous Chapters of this work have established and analysed the extent of the arbitral tribunals’ power to apply EU competition law. Quoting an expression that Voltaire, Winston Churchill, Franklin D. Roosevelt and Ben Parker have made famous through centuries, ‘With great power, comes great responsibility’. This Chapter discusses the arbitral tribunals’ duty to apply EU competition law. In light of the private nature of international arbitration, private enforcement of competition law could be further put at risk. The parties could use arbitration to enforce anticompetitive contracts and to make a cartel more stable. Hypothetically, with the complicity of arbitrators antitrust agreements could remain outside of the radar of any competition authority. Similar concerns must have influenced the Court of Justice of the European Union (hereinafter CJEU) when it answered the question whether national courts and arbitrators should have the duty to apply European competition law by virtue of their office (usually referred to as: ex officio, sua sponte, or motu proprio), even against the will of the parties. In other words, the cornerstone of this Chapter is the question whether arbitrators should be forced to abandon the judicial passivity rule and decide ex officio to extend the scope of the parties’ dispute on

181

5. ARBITRAL TRIBUNAL’S DUTIES TO APPLY EU COMPETITION LAW

competition law issues. The CJEU case-law has been interpreted as giving a positive answer to this question. Allegedly the CJEU created a new procedural rule imposing such obligation on both national courts and international arbitrators. This Chapter is divided in two Sections. In the legal analysis, that is the first Section we criticise this interpretation of the case-law, questioning whether such a rule was really created. Even assuming that this judge-made new procedural rule exists, we point out that no attention was paid to effects produced by this rule on the private parties’ economic incentives in litigation, that are discussed in the second Section. This Chapter aims to fill such gap of the legal literature using the Law and Economics methodology.

5.1

Legal Analysis

This analysis on antitrust arbitration begins from the description of the existing case-law. We cannot examine first any specific legislative provision, because none exist on the issues discussed herein. Moreover, to start directly from the legal principles could make the discussion excessively abstract. Hence, this legal analysis focuses on the Eco Swiss case and the decisions connected to it.1

5.1.1

The Background of the Eco Swiss Case

In 1986, Benetton International NV, the Dutch subsidiary of the Italian based Benetton Group SpA, signed an eight years long trademark license agreement with Eco Swiss China Time Ltd., a Hong Kong based corporation, and with Bulova Corporation, based in the United States.2 The contract provided that Eco Swiss would manufacture, market, and distribute watches under the brand Benetton-Bulova. These two companies allowed the use of their names and Bulova Corporation was in charge of verifying that Eco Swiss would produce goods which comply with the agreed quality standards. This contract seems to be an extremely 1

Case C-126/97, Eco Swiss China Time Ltd v Benetton International NV [1999] I-3055. Dutch Supreme Court (Hoge Raad) 21 Mar. 1997, NJ 1998, 207, reported in Year Book Commercial Arbitration XXIII (1998): 180-195. 2

182

common international vertical agreement and, as such, it also contains an arbitration clause. However, the contract also included a clause aiming to produce a market division between Bulova and Eco Swiss. The provision is a clear violation of the first paragraph of Article 101 of the Treaty of Functioning of the European Union (TFEU) and a prima facie breach of the European competition law.3 Nevertheless, until 1991, the contract was duly in force among the undertakers. Then, Benetton requested the termination of the contract claiming a breach on the basis of royalties miscalculations by its counter-parties.4 According to the arbitration clause, this dispute was solved under the rules of the Netherlands Arbitration Institute (NAI).5 Benetton did not raise the so called euro-defence in the arbitration proceeding.6 Hence, it did not use the European competition law argument to claim that the contract violated the first paragraph of Article 101 TFEU and therefore it was void, under the second paragraph of Article 101 TFEU. The European competition law issue was completely disregarded both by the parties and by the arbitral tribunal. The result of the arbitration proceeding was a first partial award finding that Benetton had wrongfully terminated the contract and that the license agreement was therefore still valid and in force among the parties. The quantification of the damages to be paid by Benetton was left for the final award and eventually set at the amount of 30 millions US Dollars.7 Once the arbitral award was pronounced, the only way Benetton could avoid to pay damages was by seeking either the annulment of the arbitral award or the denial of its enforcement. Benetton attempted both ways in front of the district court of The Hague.8 Only at this point Benetton raised, as basis for its requests, the violation of European competition law. Benetton argued that the 3

Dutch Supreme Court (Hoge Raad) 21 Mar. 1997, NJ 1998, 207, reported in Year Book Commercial Arbitration XXIII (1998): 180-195. 4 Dutch Supreme Court (Hoge Raad) 21 Mar. 1997, NJ 1998, 207, reported in Year Book Commercial Arbitration XXIII (1998): 180-195. 5 Nederlands Arbitrage Instituut (www.nai-nl.org). 6 On the euro-defence see Blanke [2009] 7 Dutch Supreme Court (Hoge Raad) 21 Mar. 1997, NJ 1998, 207, reported in Year Book Commercial Arbitration XXIII (1998): 180-195. 8 The latter eventually resulting in a request for preliminary ruling from the Court of Justice of the European Union (CJEU) under Article 267 TFEU.

183

5. ARBITRAL TRIBUNAL’S DUTIES TO APPLY EU COMPETITION LAW

award should be set aside as contrary to the public policy, under Article 1065 of the Dutch Code of Civil Procedure (CCP).9 The argument for this claim is based on the fact that the award supports a contract that is apparently in breach of Article 101 TFEU. The arbitral award declared that the contract was valid and was not duly terminated by Eco Swiss. If the European competition law provision should be considered as public policy under Article 1065 CCP this means that the award will be set aside if contrary to competition law.

5.1.2

The Principle of Judicial Passivity vs.

Antitrust

Rules as Public Policy In the aforementioned scenario, there are two contrasting principles that come at stake. On the one hand the principle of judicial passivity that is at the cornerstone of the Van Schijndel case.10 On the other hand, the definition of antitrust rules as public policy provisions in the procedural sense. To clarify the framework of these conflicting positions it is worth reminding that in the Van Schijndel decision the CJEU affirmed that EU competition law ‘does not require national courts to raise of their own motion an issue concerning the breach of provisions of Community [Union] law where examination of that issue would oblige them to abandon the passive role assigned to them by going beyond the ambit of the dispute defined by the parties themselves and relying on facts and circumstances other than those on which the party with an interest in application of those provisions bases his claim’.11 In other words, the parties should autonomously define the scope of the dispute. 9

Under Art. 1065(1)(3) CCP, an award may be set aside in the event the award or the way in which it was made is contrary to public policy. Other grounds for annulment are absence of a valid arbitration agreement, improper constitution of the tribunal, failure of the tribunal to comply with its mandate, and an award that has not been properly signed or that does not contain (adequate) reasoning. For an English translation of the Dutch Arbitration Act (Book 4 of the CCP), see Pieter Sanders c.s., De Nederlandse arbitragewet. English, French and German (Deventer: Kluwer, 1987). 10 C-430/93 and C-431/93 Jeroen van Schijndel and Johannes Nicolaas Cornelis van Veen v. Stichting Pensioenfonds voor Fysiotherapeuten [1995] ECR I-4705. 11 C-430/93 and C-431/93 Jeroen van Schijndel and Johannes Nicolaas Cornelis van Veen v. Stichting Pensioenfonds voor Fysiotherapeuten [1995] ECR I-4705, par. 22

184

In light of this principle, in the Eco Swiss case neither the national courts nor the arbitral tribunal could have applied Article 101 TFEU. In fact, the competition law violation was not within the scope of the arbitration proceeding. It is worth noting that if the tribunal goes beyond the scope of the arbitration proceeding, its decision can be challenged for excess of mandate.12 The arbitrators’ excess of mandate is an autonomous ground for annulment of an arbitral award and for denial of enforcement.13 Conversely such outcome could be avoided only if competition law provisions are defined as a set of rules that by their own nature need to be applied when relevant for the case, irrespectively of the parties’ will. In other words this would mean defining EU competition rules as public policy provisions in the procedural sense. The result would be that both the national courts and the arbitral tribunals are subject to the legal duty to apply EU competition law on their own motion, so that there is a specific legal obligation to apply EU competition law ex officio. The conflict between these two rules defines the research question of this Chapter. The question is which one of the these two conflicting positions should prevail: the judicial passivity role of the adjudicator or its duty to apply competition law rules ex officio, as public policy provisions? This question was at the core of the Hoge Raad’s referral of the Eco Swisscase to the CJEU for preliminary ruling, although it was phrased and structured differently.14 However, as we will see below, in the Eco Swiss case the question was not precisely answered by the CJEU in its preliminary ruling.

5.1.3

How the Adjudicator Addressed the Eco Swiss Case

The different courts facing this issue in the Eco Swiss case took different and alternative positions. The argument employed by the adjudicators proves that 12

See Article V of the NYC and ARt 1065 of the Dutch CCP, cited above. Liebscher [2003b]. 14 Case C-126/97, Eco Swiss China Time Ltd v Benetton International NV [1999] I-3055. 13

185

5. ARBITRAL TRIBUNAL’S DUTIES TO APPLY EU COMPETITION LAW

there is a trade-off to be taken into consideration. The Court of Appeal of The Hague agreed with Benetton’s view on public policy and denied enforcement of the award. According to the court, antitrust rules are at the core of the common market. In other words, the Court of Appeal solved the conflict limiting the judicial passivity role of the adjudicator. The court argued that Article 101 TFEU can be considered a public policy norm under the Dutch Civil Procedure Code, due to the fundamental importance of antitrust rules for the internal market. Hence arbitrators have to apply them motu proprio even if this means that they have to abandon the judicial passivity role.15 The matter passed then in the hands of the Dutch Cassation court, the Hoge Raad that did not share the view of the Court of Appeal.16 The Supreme Court clarified that: ‘Under Dutch law, generally, the mere fact that the contents of an arbitral award or its enforcement result in the failure to apply a prohibitory provision of competition law will not entail a violation of public policy.’ Therefore the Dutch Supreme Court did not believe there should be an obligation for arbitrators to apply Dutch competition law rules in order to prevent that their award is set aside as contrary to public policy.17 As per the application of European competition law the Hoge Raad could not provide an answer: ‘The question arises, however, as to whether this is also the case when, as in the present case, the provision concerned is a provision of Community [Union] law.’ To clarify this point the Hoge Raad referred the case to the CJEU for a preliminary ruling under Article 267 TFEU. Before the CJEU pronounced its decision, Advocate General (AG) Saggio provides an alternative solution to the two contrasting positions.18 He starts arguing that EU law should not require arbitrators that have to rule on the performance 15

Cited above. See citation above 17 It is worth noting that a different competition law, the Act on Economic Competition of 1956 was in force in the Netherlands, at that time. This principle could not be valid for the present Dutch Competition Act 1997. 18 Opinion of the Advocate General Saggio on case C-126/97, Eco Swiss China Time Ltd v Benetton International NV [1999] I-3055. 16

186

of an agreement to verify its conformity with EU antitrust law if it implies abandoning the judicial passive role and going beyond the scope of the dispute defined by the parties. They should not rely on circumstances and facts unless a party relied on them to define its claims.19 Moreover the AG believes that in light of the principle of procedural autonomy, arbitrators’ duty to raise issues of competition law ex officio should be an issue for the national laws. To support this view, the AG adds that arbitration itself, as an alternative dispute resolution means, is strongly connected to the parties’ autonomy to determine the extent of the arbitrators’ mandate.20 However this position was not followed by the CJEU. It could be argued that the problem with this view is that it sacrifices the uniform application of EU competition law throughout the Member Sates. In 1999 the CJEU decided the Eco Swiss case.21 In respect to the position taken by the Hoge Raad, the CJEU takes the opposite view. It states that Article 101 ‘constitutes a fundamental provision which is essential for the accomplishment of the tasks entrusted to the Community and, in particular, for the functioning of the internal market’.22 Then the CJEU states in an extremely clear way that: ‘A national court to which application is made for annulment of an arbitration award must grant that application if it considers that the award in question is in fact contrary to Art 101, where its domestic rules of procedure require it to grant an application for annulment founded on failure to observe national rules of public policy’.23 Although the CJEU states that an arbitral award can be annulled if contrary to EU competition law, the CJEU does not specifically take a position on the public policy nature of antitrust law. In practice it avoids to explicitly answer 19

Opinion of the Advocate General Saggio on case C-126/97, Eco Swiss China Time Ltd v Benetton International NV [1999] I-3055, par 26. 20 Opinion of the Advocate General Saggio on case C-126/97, Eco Swiss China Time Ltd v Benetton International NV [1999] I-3055, par. 21. 21 Case C-126/97, Eco Swiss China Time Ltd v Benetton International NV [1999] I-3055. 22 Case C-126/97, Eco Swiss China Time Ltd v Benetton International NV [1999] I-3055, par 36. 23 Case C-126/97, Eco Swiss China Time Ltd v Benetton International NV [1999] I-3055, par 41.

187

5. ARBITRAL TRIBUNAL’S DUTIES TO APPLY EU COMPETITION LAW

the first question asked by the Hoge Raad when referring the case for preliminary ruling. The CJEU does not say that arbitrators have to apply ex officio Article 101 TFEU even when the conflicting parties have not raised the competition law issue and in doing so arbitrators would exceed their mandate. Instead, what the CJEU did, though the above statement, was to establish, for the first time in the European legal system, the so-called second-look review of an arbitral award. This completely new procedural rule implies that the arbitral awards’ compatibility with EU competition law should be examined by Member States national courts.24 Nevertheless, through this approach, at the level of national courts, the conflict between judicial passivity and ex officio application of competition law appears to be solved. In the phase of annulment or recognition/enforcement of an arbitral award, the second-look review is independent from the procedural behaviour of the parties in the arbitral proceedings. When national courts perform the second-look they are reviewing only the result of the arbitration proceedings, so whether the award itself is in breach of antitrust rules. However, in practice the enforcement procedure extends the scope of the dispute beyond its original borders.

5.1.4

Flaws in the Eco Swiss Case-law and the Legal Practitioners’ Solutions

The problem with the approach proposed by the CJEU is two-folded. On the one hand, it is not clear whether the national courts have only to set aside the arbitral award, or whether they also have to apply ex officio competition law to the dispute. On the other hand, this approach merely relegates the problem at the arbitration level. Without a clear statement of the adjudicators’ duty to apply competition law ex officio, the conflict between judicial passivity and the public policy nature of antitrust rules is confined in the arbitration proceedings. For instance in another hypothetical Eco Swiss-alike case, the arbitral tribunal 24

Case C-126/97, Eco Swiss China Time Ltd v Benetton International NV [1999] I-3055.

188

would have two possible courses of action: either apply EU competition law ex officio or not. In the former case the award could be set aside because they exceeded their mandate and in the latter case the award could be set aside in the second-look review for violation of public policy provisions. 5.1.4.1

National Courts’ Duty to Apply Antitrust Rules Ex Officio

In practice, commentators and legal practitioners have addressed these two problems affirming that by establishing the second-look review the CJEU intended to define antitrust rules as public policy rules in a procedural sense.25 Therefore in its ruling the CJEU has implicitly affirmed the ex officio duty for the adjudicator to apply those provisions. In 2006 the CJEU pronounced the Manfredi decision, that seems to confirm this position.26 In an obiter dictum of the Manfredi case the CJEU recalls its Eco Swiss ruling and affirms that Articles 101 and 102 ’are a matter of public policy which must be automatically applied by national courts’.27 However, other legal practitioners have argued that this obiter dictum does not really invoke the ex officio application of antitrust rules.28 Firstly the Manfredi case ’did not involve elements of ex officio duties of national courts or the principle of judicial passivity of national courts in view of the ambit of a dispute as determined by the parties’.29 Secondly the obiter dictum was not part of the main line of reasoning of the Manfredi case.30 Thirdly, ‘the reference to Eco Swiss is to paragraph 39 and 40. Paragraph 39 of Eco Swiss is about the 25

Komninos [2009], pp. 221-225, more specifically in p. 224. See also Landolt [2007b], p. 77. Joined Cases C-295/04 to C-298/04, Vincenzo Manfredi v Lloyd Adriatico Assicurazioni [2006] ECR I-6619; and similarly but more recently and restricted to Article 101 case C-8/08, T-Mobile Netherlands BV, and Others v Raad van Bestuur van de Nederlandse Mededingingsautoriteit [2009] ECR I-4529, par. 49. 27 Joined Cases C-295/04 to C-298/04, Vincenzo Manfredi v Lloyd Adriatico Assicurazioni [2006] ECR I-6619, par. 31. 28 Diederik de Groot in Chapter 16 of Blanke and Landolt [2010]. 29 Joined Cases C-295/04 to C-298/04, Vincenzo Manfredi v Lloyd Adriatico Assicurazioni [2006] ECR I-6619. 30 Joined Cases C-295/04 to C-298/04, Vincenzo Manfredi v Lloyd Adriatico Assicurazioni [2006] ECR I-6619. 26

189

5. ARBITRAL TRIBUNAL’S DUTIES TO APPLY EU COMPETITION LAW

elevation of Art 101 to the rank of public policy within the meaning of Article V of the NYC’.31 As it is well known, this provision allows the denial of award enforcement under certain conditions and is not about the ex officio application of mandatory rules in contrast to the principle of judicial passivity. Most certainly the CJEU’s reference is out of context. To summarise, even if the Manfredi ruling is to be interpreted as clarifying that national courts have the duty to apply ex officio antitrust rules in the face of silence or against the will of the parties, the question is still open for arbitral tribunals. Once Member State National courts have such obligation will the same obligation be automatically echoed in arbitration?

5.1.4.2

Arbitral tribunals’ Duty to Apply Antitrust Rules Ex Officio

Extending the ex officio application to arbitrators is not as immediate as one could think. To be able to go against the will and the expectations of the parties, international arbitrators need a solid legal justification. One of the proposed approaches, to adequately justify the ex officio application of EU competition law in international arbitration, is the use of private international law.32 After all, this approach provides a straightforward solution for national courts: once the lawmaker has clarified that a set of norms are of mandatory application, then the judge will simply apply them. The national judge is subject to the international law of its forum. The forum’s mandatory rules in the international sense override the conflict of laws rules, along with the choice of applicable law made by the parties of an international contract and can even override the judicial passivity role of the judge. However, private international law was specifically created for national courts and not for international arbitrations. In the words of Pierre Mayer, ‘the relationship linking the arbitrator to the law is much more complex than the relationship 31

Joined Cases C-295/04 to C-298/04, Vincenzo Manfredi v Lloyd Adriatico Assicurazioni [2006] ECR I-6619. 32 Mayer [2001].

190

that ties judges to it’.33 In case of international arbitration, as two authors explain it, the effect of mandatory rules in arbitration is ‘one of the most difficult issues’ (Marc Blessing) ‘comporting an unfortunate degree of complexity’ (Pierre Mayer).34 An arbitral tribunal does not have a forum in the sense of a national court. For instance, the first issue that an arbitral tribunal would have to address is which set of mandatory rules to apply, having to choose among the lex causae, the lex fori, or those of a third jurisdiction such as the lex loci solutionis or the law of the place of enforcement of the award. Although the choice of law in international arbitration has been covered in a previous Chapter of this research, it should be remembered that even within different European Union Member States, the legal systems present considerable differences in this regard.35 Hence arbitral tribunals do not have an unique rule that allows them to promptly justify the application of a specific set of mandatory rules. They are expected to evaluate on a case-by-case basis and then determine the most appropriate mandatory rules in an international sense that they choose to apply to the arbitration proceedings. As Blessing explains it, arbitrators have to determine a rule’s application worthiness under a rule of reason, which adequately encapsulates most contemporary theories.36 Moreover, even when arbitrators determine the most appropriate mandatory rules to be applied, this inconclusiveness of the private international law approach supports their reasoning only insofar as the arbitration is in the context of a specific national forum. The fact that their reasoning cannot hold true, when the award is detached from the specific forum, makes the international law approach ill-suited to justify the ex officio application of EU antitrust rules. Many commentators have noted that a different path can be used to extend on arbitrators the obligation to apply antitrust law motu proprio.37 This argument is based on two elements: firstly, on the other well-established arbitrators’ 33

Mayer [2001], p. 246. Blessing [1997] p. 38; Mayer [1986], p. 292. 35 See above the Chapter on Jurisdiction. 36 Blessing [1999b], pp. 61-65; Landolt [2006] pp. 129-139 and 156-159. 37 Blanke and Nazzini [2008a], Van Houtte [2008], Komninos [2009]. 34

191

5. ARBITRAL TRIBUNAL’S DUTIES TO APPLY EU COMPETITION LAW

obligations and, secondly, on the second-look review system established by the CJEU in the Eco Swiss case. Within the list of arbitrators’ undeniable duties we can find the following: • the duty to treat the parties equally and to render a valid award; • the duty to make the best effort to reach an enforceable award;38 These two obligations are indisputably recognised and can be referred to altogether as the arbitrators’ duty to provide an healthy arbitral award.39 The second-look review, as defined in the Eco Swiss case, establishes a treat on the validity of arbitral awards.40 In light of the second-look review, a EU Member State national court can set aside or deny enforceability of an arbitral award that does not comply with the EU competition law.41 Arbitrators would not be able to comply with their duty to provide a healthy award unless they do not apply antitrust rules motu proprio. For this reason some authors have referred to a de facto duty or to an indirect duty for arbitrators to apply antitrust rules ex officio.42 The success of this argument is based on its simplicity and linearity. However, also this approach is not immune from criticism and it may not always work flawlessly. The first weak point is the fact that the healthy award duty is only a best efforts commitment. International arbitrators are not required to do whatever is in their power to fully comply with it. So, it is not a legal obligation in the strict sense. Considering that an international case can be linked to many jurisdictions, it will be unreasonable to pretend that arbitrators produce an award which complies with all mandatory rules that can come at stake. Otherwise in order to produce a healthy award, for instance in the Eco Swiss case, the tribunal 38

This duty is mitigated by the question whether arbitrators should be asked to asses the enforceability of an award under all potential jurisdictions of enforcement of the award 39 Liebscher [2003b]. 40 Case C-126/97, Eco Swiss China Time Ltd v Benetton International NV [1999] I-3055. 41 Case C-126/97, Eco Swiss China Time Ltd v Benetton International NV [1999] I-3055. 42 Blanke and Nazzini [2008a]; Komninos [2009], pp. 370-371.

192

would have to comply with the national law requirements of the Netherlands, the United States, and Hong Kong. Even assuming that these three legal systems do not have incompatible mandatory rules and that arbitrators are familiar with all of them, it is clear that this task can become extremely complex. For example, what if all the assets of Benetton are located in Italy, where are its headquarters? Then to have an enforceable award, arbitrators will have to comply with the Italian requirements. What if Bulova is an off-shore company, only incorporated in Hong-Kong for fiscal reasons while all its assets are located in Russia. Clearly the healthy award duty cannot go beyond a best efforts commitment. Moreover, assuming that Benetton has some assets outside of the jurisdiction of the European Union, the original award could be enforceable, even in light of the European second-look review. For a non European national court, an award that does not apply EU competition law could be perfectly healthy. Therefore, in order to provide an healthy award, the arbitral tribunal will comply with international public policy requirements and the domestic ones of at least one jurisdiction where the winning party would have an interest to enforce the award. A second source of criticism is based on the fact that arbitration is a truly international dispute resolution tool. It is unreasonable to expect from international arbitrators to be familiar with the requirements of all the possible second-look review jurisdictions. Especially when the scope of the dispute, as defined by the parties, does not involve such provisions. This argument is even more relevant in the field of antitrust arbitration which is not based on legislation but on case-law, hence less easily available to international legal practitioners. Moreover international arbitrators have also to choose which mandatory rules to take into account, when providing an healthy arbitral award. Notwithstanding the requirement to apply European competition law allegedly set by the Eco Swiss case, the arbitral tribunal may or may not have included EU law within their selection of mandatory rules. To summarise, considerable problems emerge from this analysis of the caselaw and the legal literature. On the one hand, both the Eco Swiss case and the Manfredi case fail to clarify whether national courts have to abandon the judicial

193

5. ARBITRAL TRIBUNAL’S DUTIES TO APPLY EU COMPETITION LAW

passivity rule in favour of the ex officio application of antitrust rules. On the other hand, it is still an open question, whether arbitrators have to give preference to the principle of judicial passivity or whether they are also under the duty to apply EU competition law motu proprio. From a legal point of view, even assuming such obligation exists for national courts, both legal approaches proposed by the practitioners do not provide a convincing justification for extending the ex officio duty from national courts to arbitral tribunals.

5.2

Economic Analysis

The legal analysis above, explains how the CJEU dealt with the conflict between the judicial passivity role and the duty to apply antitrust rules ex officio. In practice the CJEU, supported by the majority of legal practitioners, tried to establish a new legal procedural rule under which the latter of the two would prevail. Above, we have also shown some shortcomings of the approach followed by the CJEU and by the legal practitioners to introduce this new procedural rule. However the most striking shortcoming is that both the CJEU and legal practitioners have tried to create a new legal rule without taking into consideration its effects on private parties incentives. There is no argument that examines the effects that this new procedural rule would have either on parties economic incentives or on the private enforcement of competition law. This part of the Chapter provides the Law and Economics insights necessary to fill this gap of the legal literature. The goal is to verify whether the rule created by the CJEU’s case-law was the optimal solution in light of its effects on private parties incentives to bring claims to suit.

5.2.1

Background Hypothesis

We start the analysis trying to retrace the intuition behind a possible reasoning that could have supported the the CJEU’s preference for the ex officio duty, when pronouncing the Eco Swiss and the Manfredi decisions. The implicit assumption is that anticompetitive contracts decrease social welfare and are a cost

194

for the society. They impair the competition in the market; lead to an increase in prices for goods and services; lead to a decrease in the quantity and quality of goods and services exchanged in the market; and ultimately impair economic growth. Anticompetitive contracts exist because they provide private benefits to the members of a cartel or to a company with a dominant position in the market. This scenario essentially portraits the negative externalities created by anticompetitive contracts. Hence, competition law enforcement aims to insulate existing anticompetitive contracts from the market and prevent the formation of new ones. The European Commissioner for Competition, Joaqun Almunia, recently discussed this topic at the core of his speech on: ‘Competition Policy Enforcement as a Driver for Growth’.43 Competition law enforcement is pursued through two instruments. On the one hand, the public enforcement of competition law, which is based on the activity of the European Commission and of Member States’ National Competition Authorities. The society bears all the costs of the public enforcement activity. On the other hand, the private enforcement of competition law entitles private parties to pursue through litigation the cancellation of an anticompetitive contract and to seek damages from the party that has breached the EU competition law. Although the costs of litigation are usually shared between the private parties and the state, in light of the burden of proof rules, we could argue that proving a competition law violation in private enforcement is mostly a private cost.44 Moreover, as well as public enforcement, also private enforcement of competition law creates the social benefit of insulating anticompetitive contracts from the market and preventing the formation of new ones. Hence the private enforcement of competition law can be said to produce positive externalities on the social welfare. This scenario holds true, and is valid, also when the parties have chosen to solve their antitrust dispute in arbitration instead of traditional litigation. 43

Speech delivered during the Brugel workshop in Brussels on 18th February 2014, available online at:http://www.bruegel.org/nc/blog/detail/article/ 1260-competition-policy-enforcement-as-a-driver-for-growth/ 44 In arbitration, the private enforcement costs are almost exclusively paid by the private parties.

195

5. ARBITRAL TRIBUNAL’S DUTIES TO APPLY EU COMPETITION LAW

However, arbitration could create also certain risks for the antitrust private enforcement, due to the specific characteristics of this alternative dispute resolution tool. Arbitration is performed through private proceedings and the arbitrators composing the tribunal are private adjudicators selected by the parties. Hence, the legal system cannot thoroughly control every part of a dispute that is solved through arbitration. This is true, unless one of the parties decides to enforce the arbitral award with the help of a states’ ius imperium.45 It is because of this characteristic that, as explained in the legal analysis above, in arbitration, if the losing party has assets outside the EU, the arbitral award may be enforced there and never fall under the scrutiny of a European national court. In light of the potential lack of state control, it is easy to make an argument against the principle of judicial passivity in arbitration. Without the arbitrators’ duty to raise competition law issues ex officio, the parties could benefit from the use of arbitration to enforce anticompetitive agreements. In the worst case scenario we could think of arbitration as a tool to solve the disputes among the members of a cartel. Hypothetically, the judicial passivity principle in arbitration could guarantee the survival and the solidity of a cartel. The result would be even higher social costs from more stable anticompetitive agreements. Therefore, the principle of judicial passivity is apparently incompatible with the private enforcement of competition law in arbitration. On the contrary, the rule creating the duty for arbitrators to apply EU antitrust rules ex officio apparently reinforces the social benefits of the private enforcement in arbitration. Firstly, the hypothetical risk of using arbitration to enforce anticompetitive agreements vanishes completely. Arbitrators and national judges become watchdogs ready to block any anticompetitive contract. Secondly, since in arbitration the costs of solving a dispute are exclusively paid by the parties, the arbitration private enforcement comes at almost zero costs for the state. Moreover this rule seems to enhance the legal certainty and to be in the best interest of the due administration of justice. It is certainly redundant to litigate a contract once and then to re-litigate the same contract anew upon the issue 45

The sovereignty of the State.

196

whether, in the first place, the contract was null and void due to a breach of EU competition law. This certainly appears as a useless multiplication of litigation costs. Therefore, under this line of reasoning, a rule creating the duty for arbitrators to apply antitrust rules motu proprio is preferable to the alternative rule supporting the principle of judicial passivity. Assuming that the CJEU has gone through this line of reasoning, it could apparently justify the case-law of the CJEU, although no such analysis exists in the motivation of the CJEU. However, the flaw of this argument is that its primary concern is the social welfare point of view. This leads to overlook the private parties perspective. The fundamental rationale behind the principle of judicial passivity is that it guarantees the parties’ autonomy to determine the scope of a dispute. It protects the private parties’ ex ante legitimate expectations to have a dispute confined within the borders of their claims. Certainly there are specific exceptions to this principle. The most obvious exception, common to most jurisdictions, is the adjudicators duty to guarantee the application, even ex officio, of the procedural due process rules. The justification for such exception is that these procedural rules are of immediate application and guarantee the fundamental fairness of the procedure. A violation can be promptly recognised by the adjudicator. All the facts necessary to apply them can be easily accessible to the adjudicator and verified directly by him. On the contrary the application of Article 101 and 102 TFEU is a fairly complex fact-intensive exercise. European competition rules have an economic nature and their application can require an extensive economic analysis.46 The straightforward example of the economic nature of competition law rules comes directly from the structure of Article 101. A contract may contain a provision that appears to be in breach of the first paragraph of Article 101 TFEU. Nonetheless 46

See 2008 Commission Staff Working Paper, point 65.

197

5. ARBITRAL TRIBUNAL’S DUTIES TO APPLY EU COMPETITION LAW

an economic analysis can prove that the contract meets the requirements of the third paragraph of Article 101 TFUE and is therefore perfectly valid. A rule that obliges the adjudicator to apply EU competition law ex officio is de facto forcing private parties to debate complex competition law issues. Therefore, under the ex officio rule, private parties will have to bear the additional costs of a dispute that has been extended beyond its scope by the adjudicator. If the exceptions to the judicial passivity role are extended to arbitration, the private parties will have to bear the risk of having to afford these costs for all jurisdictions involved in the case that have decided to require the ex officio application of certain rules. To summarise, because of the economic nature of competition law, the rules created by the case-law of the CJEU may not provide a justified exception to the principle of judicial passivity. Moreover they can have undesired extensive adverse effects on private parties economic incentives in litigation and arbitration. Therefore the conflict between the principle of judicial passivity and the duty to apply competition law ex officio is not as simple as it can appear at first sight.

5.2.2

The Model

In this analysis we propose the use of the Shavell’s model on alternative dispute resolution to examine the adverse effects of the judge made rule discussed above on private parties’ economic incentives.47 To facilitate the reader who is already familiar with this model we will use the same notation and the same assumptions used by Shavell.48 More specifically, according to the simplest model of litigation: the parties can be considered risk neutral; each party will pay its own legal costs; the adjudicator can award the full value of the dispute at stake; the parties disagree on the merits of the case, so on the probability that the Plaintiff will win the case. 47 48

Shavell [1995]. Shavell [1995].

198

In arbitration the parties can decide how the legal expenses of the dispute shuold be allocated. Under the so-called American rule each party pays its own legal costs regardless of the outcome of the dispute. On the contrary under the so-called English rule the party that loses the dispute has to pay both its own and the winning party’s expenses. In this Chapter, the economic analysis focuses on the so-called American rule for the allocation of legal expenses among the parties. This choice is not too restrictive. Appendix 2, below, clarifies that the results of this analysis are consistent even shifting from one rule to the other. The use of the latter rule increases only the complexity of the model, whereas the resulsts are affected only in a quantitative manner.

In this scenario, let the amount at stake in the dispute be equal to x. Each party will have its own belief about Plaintiff’s probability to win the case: pπ for Plaintiff and pδ for Defendant. The costs of the trial are equal to tπ for Plaintiff and tδ for Defendant. Whereas, in arbitration, Plaintiff believes that the probability that he will win the dispute is equal to qπ , while Defendant believes that Plaintiff can win the arbitration with probability qδ . The costs of the arbitration proceeding will be aπ for Plaintiff and aδ for Defendant.49 Considering that the parties can also settle the dispute, the amount of settlement would be called s.

5.2.2.1

The Standard Model of Litigation

Starting from the structure of the simplest standard model of litigation described by Shavell and graphically represented in Figure 1 below:

49

This notation is the same adopted by Shavell. He uses the letter p for litigation in front of a national court and q for arbitration. As will be seen below the benefit of this double notation lies on the fact that it allows the reader to easily follow the model and rapidly identify the decision maker as well as the similarities between the two adjudication systems.

199

5. ARBITRAL TRIBUNAL’S DUTIES TO APPLY EU COMPETITION LAW

plaintiff

suit

settlement

no suit

trial

Once a dispute arises between Plaintiff π and Defendant δ, Plaintiff have to decide first whether to bring suit or not. Then the parties can settle the dispute for the value s or enter the trial procedure. Plaintiff will bring suit only if his expected value from trial, called EπT , is positive.50 Provided that:

EπT = pπ x − tπ

(5.1)

The participation constraint that needs to be satisfied for Plaintiff even to bring suit is that Equation 5.1 has to be higher than zero.51 This is the condition under the principle of judicial passivity. The question now is what happens to the participation constraint if a competition law issue, which is outside of the scope of the dispute, is raised by the adjudicator ex officio. The parties will be forced to litigate also on this additional issue. This means that, everything else equal, under a procedural rule that 50

In line with the notation followed by Shavell, E stands for expected value, the superscript used below T, S, B, N refers respectively to the player’s actions ‘trial, settlement, binding Arbitration and nonbinding Arbitration’ the meaning of the last two will be explained below. As it is clear at this point the subscripts either π or δ refer to which party’s perspective is into consideration. 51 In line with the setting of the Shavell model, the possibility of settlement will be discussed below.

200

obliges the adjudicator to apply antitrust rules ex officio, Plaintiff will have to pay higher t0π costs for litigating the additional antitrust issue and will also face a decreased probability p0π to win the dispute. The probability p0π that includes within the dispute the litigation of competition law, will always be smaller than pπ . An anticompetitive conduct implies that there will be a party that has the right and the incentive to censure it. When no party raises this issue, it means that they don’t believe there is such a breach of competition law. However there is always even a smallest probability that the adjudicator would find a false positive breach of competition law. Therefore, the adjudicator’s ex officio duty to apply competition law, even against the will of the parties, results in Eπ0T < EπT . Hence, it restricts the participation constraint making it more difficult for Plaintiff to bring suit. Once Plaintiff brought a suit, he would settle the dispute if Equation 5.1 is lower or equal to the settlement amount s, otherwise Plaintiff will start trial.

pπ x − tπ <= s

Conversely Defendant will settle only for an amount lower than or equal to his expected cost of litigation, that are called EδT :

s <= pδ x + tδ

As illustrated also by Shavell this means that the parties will settle when both conditions are verified simultaneously:

pπ x − tπ <= s <= pδ x + tδ

201

5. ARBITRAL TRIBUNAL’S DUTIES TO APPLY EU COMPETITION LAW

This condition can also be equivalently expressed in the following form:

(pπ − pδ )x ≤ tπ + tδ

(5.2)

Provided that under the ex officio application of antitrust rules the litigation costs will increase for both parties (right side of Equation 5.2), this rule makes it more likely that Equation 5.2 is verified, so that the parties will settle more often. One should keep in mind that settlements are usually reached out of the court, hence even if the contract is anticompetitive a settlement will not remove it from the market. Proposition 1: in ordinary litigation, if judges have the duty to apply antitrust rules ex officio, the parties will bring less suits. Even when the Plaintiff brings a suit, the private enforcement of competition law will decrease because parties will settle more often.

5.2.2.2

The Model With Arbitration

At this point we can also consider how the CJEU’s case-law changes what will happen according to Shavell’s model, including among the choices of the parties international arbitration. As described by Figure 5.2.2.2 below,52 along with the options available in Figure 1, now we can compare the Plaintiff incentives under two additional regimes, namely the binding arbitration and nonbinding arbitration.

52

Taken from Shavell [1995], p. 17.

202

Plaintiff

suit

settlement

binding arbitration

no suit

trial

nonbinding arbitration

settlement

trial

To clarify this notation, we can think that the binding arbitration option corresponds to a scenario where the arbitral award can be directly enforced by the winning party. While the nonbinding arbitration is one that to be enforced needs to pass through the second-look review procedure established by the CJEU in the Eco Swiss case, discussed at the beginning of this Chapter. To compare and rank the different actions available to the parties, Shavell calculates the net expected value of each action A as: ∆E A = EπA − EδA The action A with the higher net expected value ∆E A is Pareto superior to another action A0 when ∆E A > ∆E 0A , so when its net expected value is superior. Therefore using Shavell’s model, we can understand how rational parties incentives will be affected by the studied rules. Starting from the easiest option, if the parties settle their ∆E S = s − s = 0, as it is obvious, since Plaintiff receives what is paid by Defendant. Whereas, if the parties choose to go directly to trial the net expected value will be equal to:

EπT − EδT = (pπ − pδ )x − (tπ + tδ )

203

(5.3)

5. ARBITRAL TRIBUNAL’S DUTIES TO APPLY EU COMPETITION LAW

This result is consistent with the previous part of this model. In fact, according to what is just stated, the parties will settle under the condition that the net expected value of trial ∆E T is less than the net expected value of settlement ∆E S . This is the case if Equation 5.3 is lower or equal to 0. This is exactly the same requirement defined in the previous Section as settlement constraint.53 Similarly to what Shavell does in his model, the next two actions, namely binding and nonbinding arbitration, will be examined under two different scenarios. On the one hand we will discuss the case when arbitration proceedings perfectly predicts the trial outcomes. In our analysis this will happen if arbitral tribunals are subject to the same rule as national courts, namely either both of them are subject to the duty to apply competition law ex officio, or both of them are not under such obligation. On the other hand, we have the case when arbitration proceedings do not predict the trial outcomes. This happens when the two adjudicators are subject to different rules; in other words, the arbitral tribunal applies the principle of judicial passivity, while EU national courts do not. 5.2.2.3

When Also Arbitral Tribunals Apply Antitrust Rules Ex Officio

When the ex officio application of competition law in arbitration perfectly predicts whether also trial would result in antitrust ex officio application, we could think to this case as qπ = pπ , then the net expected value of binding arbitration is equal to:54

EπB − EδB = (pπ − pδ )x − (aπ + aδ )

53 54

See Equation 5.2. Shavell’s A10 Equation, Shavell [1995], p. 24.

204

(5.4)

Obviously, everything else equal, binding arbitration will be preferred to trial if (aπ + aδ ) < (tπ + tδ ). That is to say if the costs of binding arbitration proceedings are less than that of trial. As per nonbinding arbitration, remembering that now we are discussing the case when arbitration is perfectly predictive value of trial, if Plaintiff loses in arbitration the parties will not go to trial. Whereas, if Plaintiff wins in the nonbinding arbitration, the parties will not go to trial for the second-look review, since it will only increase their litigation costs, but will settle for an amount x − tπ < s < x + tδ . Therefore, the net expected value of nonbinding arbitration is equal to:55

EπN − EδN = (pπ − pδ )s − (aπ + aδ )

(5.5)

In light of the fact that the parties will settle for an amount s, the value of the dispute becomes this amount, instead of x that was the original value. Hence, nonbinding arbitration will be preferred to binding arbitration only in the unlikely case that the settlement amount s is superior to the original value of the dispute x. We can now summarise what this means for the issue of our interest in this Chapter. Firstly, if both arbitral tribunals and national courts have the duty to apply antitrust rules ex officio, private parties will prefer arbitration. We implicitly assume that for international business parties, international arbitration is less expensive, namely (aπ + aδ ) < (tπ + tδ ) is verified. Therefore, arbitration decreases the participation constraint of trials. Plaintiff will start arbitration proceedings for some cases that he would have never brought to trial. Everything else equal, if both arbitrators and judges follow the rule of ex officio duty, the private enforcement of competition law will increase with arbitration, as compared to litigation. Secondly, the second-look review created by the CJEU converts the 55

Shavell’s Equation A13, Shavell [1995], p. 25

205

5. ARBITRAL TRIBUNAL’S DUTIES TO APPLY EU COMPETITION LAW

arbitral award that needs to be enforced in Europe into nonbinding arbitration, hence it relegates the winning parties under a Pareto inferior option unless the settlement amount s is higher than the value of the dispute x. Furthermore, under the second-look review system, when s < x fewer cases will satisfy the participation constraint requirement necessary for Plaintiff to bring a suit, in comparison to a scenario without the second-look review. Moreover, given that the arbitrators follow the duty to apply antitrust rules ex officio making arbitration predictive of the second-look review, the parties will settle and will not go to litigation. Therefore the primary goal of the second-look review, which was a state control of the arbitrators’ conduct, is failed. Proposition 2: Assuming that arbitrators comply with the national courts’ rule to apply antitrust rules ex officio, the CJEU’s Eco Swiss decision that created the second-look review, produce adverse effects on the private parties incentives to litigate disputes. 5.2.2.4

When Arbitral Tribunals Refuse to Apply Antitrust Rules Ex Officio

In the context of this work an arbitration that has no predictive value of trial outcomes, means that the arbitrators do not apply competition law ex officio and give prevalence to the principle of judicial passivity. In this scenario the net expected value of binding arbitration is equal to:56

EπB − EδB = (qπ − qδ )x − (aπ + aδ )

(5.6)

However the problem is that in the context of this analysis, binding arbitration is restricted to those cases in which the losing party has assets outside of the EU, and the winning party succeeds to enforce the award without incurring into the second-look review of the European national courts. 56

Shavell’s Equation A16, Shavell [1995], p. 25.

206

The only arbitration option viable for those parties that will not be able to enforce the arbitral award without the EU national courts’ second-look is nonbinding arbitration. When arbitrators do not apply antitrust rules ex officio, after nonbinding arbitration the parties will be in the same position as if there has been no arbitration, with respect to the competition law issue. Provided that arbitrators will address the scope of the dispute as defined by the parties, the parties’ assessment of Plaintiff’s probability to win arbitration are equal to qπ and qδ . Then, in light of the initial assumption of this analysis, the judge in the subsequent trial will address only the competition law issue. Hence, competition law issues are out of the scope of arbitration and the parties will estimate Plaintiff’s probabilities to win the second-look trial a equal to p0π and p0δ . As explained above, these two disputes are independent events. Hence, in the views of the parties, the probability that Plaintiff will win both nonbinding arbitration and the subsequent second-look trial, will be respectively equal to qπ p0π and to qδ p0δ . In light of the fact that the parties will determine these last two probabilities ex ante, there is no reason to suspect that their value will be different from the value of the probabilities that the parties will attach to Plaintiff’s victory at direct trial. In other words, the Plaintiff’s probability to win the whole procedure (nonbinding arbitration and subsequent trial on antitrust rules) will be equal to his probability of winning the whole case at direct trial without passing through arbitration at all. This means that we can set qπ p0π = pπ and to qδ p0δ = pδ . At this point we can calculate the net expected value of the nonbinding arbitration as equal to:

EπN − EδN = (pπ − pδ )x − (aπ + aδ ) − (tπ + tδ )

(5.7)

The net expected value of direct trial will be the same as in Equation 5.3, namely:

207

5. ARBITRAL TRIBUNAL’S DUTIES TO APPLY EU COMPETITION LAW

EπT − EδT = (pπ − pδ )x − (tπ + tδ )

From the above it is clear that the net expected value of the direct trial is Pareto superior to the net expected value of nonbinding arbitration. Hence, direct trial will be preferred by the parties. Nonbinding arbitration will result merely in higher costs for the parties to solve the dispute, in comparison to direct trial. To summarise, in the hypothesis that arbitrators do not apply competition law rules ex officio while national courts do, arbitration will be a viable option only if the parties of the dispute have assets outside the EU. Otherwise, as described at the beginning of this Chapter, the parties will have to litigate in front of the national court the competition law issues even if they were outside of the scope of the dispute in arbitration, to be able to enforce an arbitral award in the EU. This will increase their costs providing them with a solid incentive to opt-out of arbitration, in favour of direct trial in traditional litigation. Proposition 3: Assuming that the CJEU’s case-law creates both a duty to apply competition law ex officio and a second-look review to prevent the enforcement of arbitral awards that do not comply, if arbitrators do not comply with the ex officio duty, private parties that do not have assets outside of the EU will opt-out of arbitration.

208

Chapter 6 Conclusions 6.1

The Relevance of The Proposed Legal Framework Resulting from Our Law and Economics Analysis

This conclusive Chapter summarises the entire research project and elaborates on its significance. The goal of this work was to outline a legal framework that is capable to strengthen the private enforcement of EU competition law. More specifically, the policy proposals provided herein focus on the procedural law aspects of private antitrust enforcement in arbitration. This topic is permeated by legal uncertainty and has not yet fallen under the scrutiny of the Law and Economics debate. This work’s policy proposals are based on the functional approach to Law and Economics. Such research methodology uses the economic analysis of law to study how contrasting legal solutions affect private parties’ economic incentives; and it can promote a more qualified decision making process by: adjudicators, private parties and lawmakers. The chosen methodology allows us to identify failures and inefficiencies of legal provisions that are available in the market of legal rules. Therefore, the ultimate result obtained applying this methodology to antitrust arbitration is a legal framework that can increase the deterrent effect of EU competition law and the private parties’ compliance with it. The resulting framework of procedural rules would be a cost-effective policy

209

6. CONCLUSIONS

tool that could sustain the European Commission’s effort to guarantee a workable level of competition in the EU internal market. This project aims to answer the following research question: which procedural rules can improve the efficiency of antitrust arbitration by decreasing litigation costs for private parties on the one hand, and by increasing private parties’ compliance with competition law on the other hand? Throughout this research project, this broad question has been developed into research sub-questions revolving around several key legal issues that have fallen under our scrutiny. The chosen research sub-questions, summarised below, define the core of this work and result from the unresolved legal issues that do not find a solution in national legislations. This legislative gap in antitrust arbitration has stimulated the legal practitioners’ debate, producing a flow of contrasting solutions based on often inconsistent arguments. The legal framework proposed in this research project could prevent such a blurry scenario from impairing the EU private enforcement of competition law in arbitration. Therefore, our attention was triggered by those legal issues whose proposed solutions lead to relevant uncertainties and that are most suitable for a Law and Economics analysis.

6.2

The Arbitrability of Competition Law Disputes

The antitrust arbitrability question is a preliminary issue that needs to be clarified before any other legal issue is discussed. The arbitrability issue requires to determine whether a legal system considers arbitral tribunals as a “suitable forum” where private disputes based on competition law rights can be solved. The Law and Economics analysis of this work allows us to argue in favour of the wide acceptance of antitrust arbitration across jurisdictions. The critics of antitrust arbitration are afraid that private parties may use this alternative dispute resolution tool to limit the application of competition law

210

rights. The main concern is the fact that arbitration is an international private proceeding based on contractual arbitration clauses. This means that in theory a party that has a relevant market power has also the incentive to use its bargaining power to impose on the weaker party an arbitration clause that circumvents the application of competition law. Arbitration clauses, negotiated ex ante and including both a choice of law and a choice of arbitration seat provisions, can be drafted either in a pro-competitive or an anticompetitive flavour. An anticompetitive arbitration clause could make it extremely difficult or practically impossible for the weaker party to enforce its competition law rights. In such circumstances, the stronger party will be able to breach competition law with impunity. Moreover, the weaker party may not even be able to balance the loss of its antitrust rights raising the price of the contract or using other negotiated contractual clauses. The reasons for this could be, for instance, asymmetric information between the parties about antitrust issues or the dominant position of the counterparty in the market. This does not mean that the weaker party will sign a contract that is economically unprofitable (although a vertical agreement can become economically unprofitable into a dynamic perspective). The anticompetitive contract may simply provide a lower profit for the weaker party, rather than the one with a pro-competitive contract. In fact, it is rational for the weaker party to accept disadvantageous clauses as long as, in toto, the contract is mutually beneficial for both parties. Nevertheless, such anticompetitive contracts certainly do not favour social welfare nor the level of competition in the market and cannot be tolerated within a modern legal order. Therefore, the challenge faced by the lawmaker is to limit the risk that private parties use antitrust arbitration as a tool to enforce anticompetitive contracts. The legal system cannot tolerate that one party is forced to renounce to its competition law rights, neither it can enforce arbitration clauses or arbitral awards that circumvent competition law. The simplest and initially preferred solution was to totally deny the arbitrability of competition law matters. However the historic legal evolution on this topic led legal practitioners to recognise the advantages of antitrust arbitration and to gradually change their attitude. In this work we have used game theory

211

6. CONCLUSIONS

to present Law and Economics arguments in favour of the wide arbitrability of antitrust matters. Our analysis starts showing why private parties should be allowed to enforce their competition law rights. In fact, our game theory analysis shows that they may have the incentive to act as widespread guardians of the integrity of competition in a market. On this premise, we have used game theory to explain that, although the lack of state control over international arbitration proceedings can raise competition law concerns, in the context of international business, private enforcement in arbitration has also the potential to solve the shortcomings of private actions for damages (in front of national courts). Therefore our thesis is that arbitrability of antitrust matters should not be limited but widely recognised in light of its advantages for the competition law enforcement into a global economy. Meanwhile, the concerns that antitrust arbitration raises can be tackled, or at least mitigated, using appropriate legal rules. These control tools, however, need to be widely accepted and to be uniform across jurisdictions.

6.3

Pro-competitive and Anticompetitive Arbitration Clauses

As a matter of fact, currently, we are not aware of any jurisdiction that explicitly denies in toto the arbitrability of competition law matters. However, there are limitations and safeguards used to address the concerns raised by antitrust arbitration. These control tools can be various, contrasting and incompatible across jurisdictions. The lack of uniformity creates detrimental pockets of uncertainty in antitrust disputes characterised by elements of internationality. The procedural rules used by legal systems to control antitrust arbitration (the antitrust arbitration control tool) can be divided in two categories. Those that intervene ex ante at the jurisdictional level and those that intervene ex post during the so-called second-look review of an arbitral award. The former aims to prevent only issues of objective and subjective arbitrability mainly concerning the arbitration clause. The latter aims to validate the whole arbitration proceeding

212

since it can stop infringements both in the arbitration clause and in the conduct of the arbitrators, as emerging from the arbitral award. In fact arbitration proceedings are not within the control of a national legal system unless one of the parties brings the award before a national judge asking its recognition, enforcement or its annulment. This Section and the next one review our conclusions on the jurisdictional issues affecting the validity of an arbitration clause, while the last two Sections cover our findings in regards to the conduct of the arbitrators when addressing issues on the merits of the case. Starting from the question whether an arbitration clause will be declared partially or totally void for a competition law violation. When the parties draft an arbitration clause it can be shaped either in a pro-antitrust enforcement fashion or in such a way to make it practically difficult or virtually impossible to enforce competition law based rights. However different jurisdictions and different adjudicators adopt an extensive or a limited standard of control when assessing the validity of arbitration clauses. What will be considered a valid clause in certain jurisdictions is likely to be set aside in others during the second-look review. In fact, the inconsistent and uncertain judge-made rules on antitrust arbitration imply that the existing limitations to subjective and objective arbitrability widely depend on two elements: (i) who is the adjudicator entitled to decide on the issue of arbitrability; and (ii) the conflict of law rule that this adjudicator used to decide the applicable law to the question of arbitrability (either objective or subjective arbitrability). The lack of uniformity in the control tool of arbitration clauses leaves room for the parties’ strategic behaviour. However, in light of the fundamental importance of competition law for certain legal systems, more restrictive jurisdictions struggle to tolerate the more liberal approach towards antitrust arbitration of other jurisdictions. Meanwhile, the wide and smooth circulation of arbitral awards, so essential for international trade, suffers the consequences of this conflict. This research project has extensively clarified both in a legal and in an economic analysis the failures and inefficiencies resulting from the cross-jurisdiction differences in the two elements mentioned above. It has also clarified how the

213

6. CONCLUSIONS

validity of an arbitration clause, as well as the very outcome of the dispute, can be determined by these two elements of procedural law. Therefore, we agree with those scholars who suggest solving this problem through the harmonisation of the most pervasive control tool, namely the second-look review of arbitral awards. However the long discussions and the numerous proposals of modification of the second-look review rules have been fruitless. Therefore, there are good reasons to agree with Luca Radicati di Brozolo when he states that it is time to ‘accept that international uniformity in respect to the second-look review is never going to succeed’.1 Instead, our game theory analysis suggests an alternative solution based on the pursuit of a uniform private international law rule. A widely accepted and objective choice of law rule has the potential to increase the legal certainty and to coexist with the legal systems’ inconsistencies on antitrust arbitration jurisdictional rules. The solution we propose is based on the application of the doctrine of effects. This tool, already used by antitrust authorities to determine the extraterritorial application of competition rules, is able to guarantee the international uniformity of the choice of law rule. It is based on a factual analysis of the objective economic effects of an allegedly anticompetitive contract. The application of this tool allows the parties to build ex ante reliable expectations about the law applicable both to the jurisdictional matters and to the merits of the case. Both arbitral tribunals and national courts’ discretionary power to decide the law applicable on jurisdictional issues will be limited by the economic analysis underneath the doctrine of effects and the differences in procedural rules will not be detrimental. National courts or arbitrators will determine ex ante where anticompetitive effects are produced and they will solve the jurisdictional issues on antitrust arbitrability according to the applicable antitrust rules. The resulting differences in the outcome of the choice of law issue should be reduced. Moreover, under this choice of law rule, also the ex post second-look review is unlikely to produce inconsistent results across different jurisdictions. Well argued arbitral awards based on an sound analysis of effects should not be set aside. In fact, the national courts will not have to take on themselves the task of perform1

Radicati di Brozolo, L., Chapter 22 in Blanke and Landolt [2010].

214

ing a new analysis but they will just verify the consistency of the reasoning of the arbitral award on that matter. Therefore, a choice of law rule based on the doctrine of effects improves the legal certainty and the compliance with competition law favouring also international commerce. To summarise, we propose that jurisdictional issues concerning the matter of arbitrability follow the applicable competition regime resulting from a choice of law rule based on the economic analysis of the anticompetitive effects.

6.4

Cooperation Between Competition Authorities and Arbitral Tribunals in Conflict of Law Scenarios

In the legal framework emerging from this work, the choice of law rule proposed in the previous paragraph requires that both national courts and arbitral tribunals have the investigative powers that are needed to determine where anticompetitive effects are produced. In other words, to perform correctly the economic analysis of effects, adjudicators should have the investigative powers that are necessary both to gather all relevant information and to determine in the market of which legal system anticompetitive effects are likely to be produced. Hence, another question that drew our attention is whether arbitrators shall have the same powers granted to national judges to solve antitrust disputes. Legal systems grant to national courts and to arbitral tribunals certain powers that will help them to solve both factual and legal issues. Accordingly, this work investigates the differences between the powers of these two adjudicators in respect to both elements; the powers to clarify factual issues are summarised in this Section, while the next Section discusses those liked to legal issues. Regarding both elements, the results of this work suggest that when solving disputes based on EU competition law, the arbitral tribunal must be entitled to use the same

215

6. CONCLUSIONS

powers available to national courts and necessary to comply with the enforcement standard required by the public policy rules. In respect to the doctrine of effect discussed above, neither arbitral tribunals nor national courts are the most appropriate subjects to perform the required factual analysis. Competition authorities are the subjects that have all the necessary tools to perform this fairly complex fact-intensive economic analysis based on market information. National courts of European Member States can easily request the assistance of the European Commission through the cooperation mechanism that the Regulation 1/2003 made available to them. It is entirely within the discretion of national courts to decide whether to ask for this form of assistance, or not. While the competition authority is “obliged” to provide the requested assistance. Moreover the requesting court is not formally obliged to conform its decision to the outcome of this assistance by competition authorities, for the sake of the principle of due process. However there is no legal provision that can be used as legal basis to claim that arbitral tribunals have the formal power to request the assistance of the Commission, forcing the latter to comply. The question of interest studied in this work is whether also arbitral tribunals should have the power to request the assistance from the European Commission or for that matter from any other National Competition Authority. We have discussed the advantages, disadvantages and feasibility of the extension of the arbitral tribunals’ powers to include the cooperation rules valid for national courts. The results of this analysis show that is not only possible but also desirable that arbitral tribunals indirectly use this cooperation mechanism. To use it indirectly means that, although there is no power-duty relationship between arbitrators and the Commission, the mechanism could still be used by analogy to help arbitral tribunals apply EU competition law. Nevertheless there are still questions left open for further investigations. Firstly, from the legal viewpoint, it is necessary to reconcile the private international arbitration proceedings with a cooperation tool that involves a national administrative authority (or super-national in the case of the EU). This implies that, in

216

compliance with the fundamental principle of party autonomy, it is appropriate to have the parties’ agreement on the use of this cooperation tool. Alternatively, we suggest that legal systems and arbitration institutions include a provision on this matter into their lex arbitri procedural rules. Moreover the cooperation between arbitrators and either national courts or the European Commission may raise serious confidentiality issues or doubts in respect to the principle of due process. Notwithstanding the above mentioned uncertainties, further research on this topic will be relevant. The cooperation mechanism could not only bring more certainty and uniformity on the conflict of law rules in antitrust arbitration, but it could also provide a mechanism to address public policy clashes between jurisdictions. For instance, if there are two competition law regimes potentially involved into the dispute, a game theory analysis seems to suggest that two competition authorities are better suited to address the conflict of law issue. In fact these authorities are more likely to use comity knowing that they are playing a repeated game instead of the one shot game played by national courts and arbitrators.

6.5

Limiting the Second-look Review of an Arbitral Award through the CJEU’s Preliminary Rulings in Arbitration

The previous Section argues that the arbitrators’ power to cooperate with antitrust authorities can be of fundamental importance to solve the jurisdictional issues of antitrust arbitration. However, it can also be a key element to correctly decide on the merits of the case. In fact, the arbitrators’ power to cooperate with competition authorities can be used to clarify factual issues essential for the case. Hence, that is the bridge-topic that shifts the focus of this work from issues of jurisdiction to issues of merits.

217

6. CONCLUSIONS

This Section summarises the results of our work in respect to the arbitrators’ powers that are necessary to solve legal issues which are fundamental to decide the merits of a case. With regard to adjudicators’ powers, when they are facing legal issues, the European Treaties entitle national courts to refer a legal problem to the CJEU for a preliminary ruling on the interpretation that should be given to EU law provisions. According to the current CJEU’s case-law, arbitral tribunals are not entitled to request preliminary rulings to the Court of Justice of the European Union under Article 267 TFEU. Although an arbitral tribunal can apply the EU law, the CJEU does not consider it as a court or tribunal of a Member State in the meaning of Article 267 TFEU. Nevertheless the CJEU needs to guarantee the uniform application of EU law also when it is applied by arbitral tribunals. In fact, the EU legal system allows the antitrust arbitrability counting on the fact that the involved public policy interests can be safeguarded by national courts during the second-look review of the award. This ex post control tool is employed to validate the conduct of arbitrators as emerging from the arbitral award. However, if arbitrators cannot refer cases to the CJEU for preliminary ruling, it may be difficult for them to meet the requirements of the extended second-look review. This means that when arbitrators are facing a legal issue that requires a CJEU preliminary ruling, the effectiveness of their award may be limited and may be at risk the very essence of the arbitrability principle expressed above. Arbitrators have the power to decide whether mandatory norms are applicable or not and at the same time cannot assess whether a preliminary ruling is needed. Hence, during the second-look review of the award, a national court has to examine the arbitrators’ application of European Union Law. The review can be performed in the context of any method of recourse available under the relevant national legislation and most importantly it has to inquire whether a preliminary ruling is necessary to decide the case. The question that is at the centre of this work is whether arbitral tribunals should or should not have the power to refer cases for preliminary ruling to the CJEU. In light of the methodology adopted in this work we have studied

218

how the two alternative rules affect the private parties economic incentives in litigation. Our aim was to compare, from the private parties’ perspective, which of the two procedural rules create a more costly enforcement procedure. Our economic analysis shows that parties’ rent-seeking expenditure and the total lost treasure in antitrust arbitration will be higher when arbitral tribunals’ cannot use Article 267 TFEU in comparison to when they can use it. Therefore, denying arbitral tribunals jurisdiction to refer cases for preliminary ruling risks to increase arbitration costs in comparison to national courts’ litigation costs and deprive of effectiveness an arbitral award based on EU competition law. It is our thesis that the result of the current CJEU’s case-law affects private parties’ ex ante incentives to exclude competition law from a contractual arbitration clause. On the contrary, if arbitral tribunals have the power to request preliminary rulings under Article 267 TFEU, the function of the second-look review as last gatekeeper will not be that necessary and so extensive as it is today.

6.6

Arbitrators’ Duty to Apply Antitrust Rules of Their Own Motion, a Misfire?

This research project supports the expansion of arbitral tribunal’s powers to apply EU competition law. The legal system entrusts private adjudicators with the application of mandatory norms of public policy nature, but it cannot tolerate that arbitrators overlook or neglect the application of these fundamental provisions. Hence this work has discussed the question whether national courts and arbitrators should have the duty to apply European competition law by virtue of their own motion (usually referred to as: ex officio, sua sponte, or motu proprio), even against the will of the parties. A positive answer to this question implies that arbitrators should be forced to abandon the judicial passivity rule and decide ex officio to extend the scope of the parties dispute on competition law issues. Regarding national courts, the CJEU’s case-law has been interpreted as giving a positive answer to this question.

219

6. CONCLUSIONS

Allegedly the CJEU created a new procedural rule imposing such obligation on national courts. This work criticises such interpretation of the case-law, questioning whether a similar rule was really created. Even assuming that this judge-made new procedural rule exists for national courts, the question of its existence is still open for arbitral tribunals. Moreover, we point out that no attention was paid to the effects produced by this rule on the private parties’ economic incentives both in litigation and in arbitration. Under the ex officio rule, private parties will have to bear the additional costs of a dispute that an adjudicator has extended beyond its scope. Furthermore, because of the economic nature of competition law, the ex officio rule created by the case-law of the CJEU may not be a justified exception to the principle of judicial passivity. Consequently, it can produce unexpected adverse effects on private parties economic incentives in litigation and arbitration. In the economic analysis of this matter, we adapt the Shavell’s model on alternative dispute resolution to examine the adverse effects of the ex officio rule discussed above on private parties’ economic incentives. This analysis shows several counter-intuitive unexpected effects of the ex officio rule, which go diametrically in the opposite direction of what intended by the lawmaker. Firstly, in ordinary litigation, if judges have the duty to apply antitrust rules ex officio, the parties will bring less suits. Consequently the level of private enforcement of competition law will decrease. Moreover, even when the Plaintiff brings a suit, parties will settle more often, still decreasing the number of cases decided by national courts. Secondly, if both arbitral tribunals and national courts have the duty to apply antitrust rules ex officio, private parties will prefer arbitration to ordinary litigation. This rule makes arbitration predictive of the result of the second-look review in litigation and the latter will be considered redundant. Meanwhile, the private enforcement of competition law in arbitration will increase. Moreover, our economic analysis seems to indicate that the second-look review system de-

220

creases the total number of cases brought to court, both in arbitration and in litigation. The increased costs of the enforcement procedure will lead parties to settle more often, and both private enforcement and the second-look review will lose effectiveness. Thirdly, under the alternative rule: arbitrators do not apply competition law rules ex officio while national courts do. The private parties will opt-out of arbitration. In fact, this rule requires that, in order to be able to enforce an arbitral award in the EU, the parties will have to litigate the competition law matter in front of a national court; irrespective of the fact that this matter was outside of the scope of the dispute in arbitration. This procedure increases the costs of arbitration providing the parties with a strong incentive to opt-out of arbitration in favour of direct trial by national courts. Arbitration remains a viable option only in case the parties of the dispute have assets outside the EU, where the arbitral award could be enforced independently of the ex officio application of EU competition law.

221

6. CONCLUSIONS

222

Appendix 1 We show here the extensive mathematical calculations used to reach Equation 4.5 (that you can find below in this appendix called Equation 1) describing the first order condition that allows parties to maximise their shares in the rent-seeking competition in our model. The model we create analyses parties’ net share in a three-phase rent seeking competition described as follows:

RcA = 3p2A − 2p3A − 3Ac ; ∀pA = RcB = 3p2B − 2p3B − 3Bc ; ∀pB =

Arc ; Arc +Bcr Bcr ; Arc +Bcr

3r 3A2r c c − (Ar2A r 3 − 3Ac (Arc +Bcr )2 c +Bc ) 2r 3Bc 2Bc3r RcB = (Ar +B r )2 − (Ar +B r )3 − 3Bc c c c c

→ RcA = →

In order to maximise their net share of the rent, for plaintiff in respect to A and for defendant in respect to B, parties will choose A∗c and Bc∗ that satisfy the following first order conditions:

δRcA 2rA2r−1 Bc2r c = −1=0 δAc (Arc + Bcr )4 δRcB 2rBc2r−1 A2r c −1=0 = δBc (Arc + Bcr )4

(1)

What follows are the main mathematical calculations we use to reach the first order condition above. In order to simplify notations, for the purpose of this mathematical proof, here we set Ac = A. Moreover, given that the two equations

223

. APPENDIX 1

above are symmetrical, it is sufficient to calculate only one of them (RcA ). The solution will be applicable also to the second equation (RcB ). Before dealing with the first derivative of RcA in respect to A, we first elaborate a bit more on the starting point, equation RcA itself, to simplify it as much as possible:

RcA =

3A2r 2A3r − − 3A (Ar + B r )2 (Ar + B r )3

RcA =

3A2r (Ar + B r ) − 2A3r − 3A (Ar + B r )3

RcA =

3A3r + 3A2r B r − 2A3r − 3A (Ar + B r )3

RcA =

A3r + 3A2r B r − 3A (Ar + B r )3

Now, it is easier to calculate the first derivative of RcA in respect to A and simplify it :

δRcA (3rA3r−1 + 6rB r A2r−1 )(Ar + B r )3 − (A3r + 3A2r B r )(Ar + B r )2 (3rAr−1 ) = −3 δA (Ar + B r )6 δRcA (3rA3r−1 + 6rB r A2r−1 )(Ar + B r ) − (A3r + 3A2r B r )(3rAr−1 ) −3 = δA (Ar + B r )4 δRcA 3rA4r−1 + 6rB r A3r−1 + 3rB r A3r−1 + 6rA2r−1 B 2r − 3rA4r−1 − 9rB r A3r−1 −3 = δA (Ar + B r )4 δRcA 6rA2r−1 B 2r = −3 δA (Ar + B r )4

We can now move to the first order condition of the maximisation problem.

224

Parties will choose value of A and B that satisfy the following first order condition:

δRcA 2rA2r−1 B 2r = 0; → =1 δA (Ar + B r )4 δRcB 2rB 2r−1 A2r = 0; → =1 δB (Ar + B r )4

(2)

The result of this maximisation problem are values of A and B that satisfies both the equations above. This is true for each value of A∗ = B ∗ . Therefore setting A = B we can define both these values in function of r:

∀A = B; →

r r r 2rA4r−1 = 1; → 3 = 1; → r = 8A; → A = ; B = r 4 (2A ) 2A 8 8

225

. APPENDIX 1

226

Appendix 2 This Appendix argues that even shifting from the American Rule to the English Rule of legal expenses allocation the results of the model described in Chapter 5 are consistent. The use of the latter rule increases only the complexity of the model and affects the results only in a quantitative manner. Under the American Rule, Plaintiff’s expected value from trial is equal to: EπT = pπ x − tπ

(3)

Under the English Rule, Plaintiff’s expected value from trial is equal to: EπT = pπ x − (1 − pπ )(tπ + tδ )

(4)

Respectively Defendant’s expected value from trial under the American rule is equal to: EδT = pδ x + tδ

(5)

Defendant’s expected value from trial under the English rule is equal to: EδT = pδ (x + tπ + tδ )

(6)

In the model of Chapter 5 we use Equation 3 and 5 (under the American rule) to compare and rank the different actions available to the parties calculating the net expected value of each action A as: ∆E A = EπA − EδA The action A with the higher net expected value ∆E A is Pareto superior to another action A0 when

227

. APPENDIX 2

∆E A > ∆E 0A , so when its net expected value is superior. Extending the model of Chapter 5 to the English rule would require to perform the same exercise using Equation 4 and 6. Under the English rule, when the parties settle their ∆E S = s − s = 0. When the parties choose to go directly to trial, the net expected value will be equal to: EπT − EδT = pπ x − (1 − pπ )(tπ + tδ ) − pδ (x + tπ + tδ ) After rearranging, this equation becomes: EπT − EδT = (pπ − pδ )x + (pπ − pδ )(tπ + tδ ) − (tπ + tδ )

(7)

In light of the fact that ∆E S = s − s = 0 the parties will always settle when EπT − EδT is negative, this means that Equation 7 is relevant for the model only when (pπ − pδ ) is not negative. This condition implies that 0 <= (pπ − pδ ) <= 1. Therefore, in Equation 7, (pπ − pδ )(tπ + tδ ) < (tπ + tδ ), resulting in −(tπ + tδ ) < (pπ − pδ )(tπ + tδ ) − (tπ + tδ ) < 0. Comparing this result under the English rule with the corresponding equation under the American rule, it is clear that the difference between the value of EπT − EδT under the American rule and that value under the English rule is only quantitative not qualitative. Consistently with this statement we can continue to apply the English rule to the model of Chapter 5. The net expected value of binding arbitration is equal to: EπB − EδB = (pπ − pδ )x + (pπ − pδ )(aπ + aδ ) − (aπ + aδ )

228

(8)

Comparing the net expected value of a direct trial, Equation 7, with the net expected value of binding arbitration, Equation 8, we reach the same result of Chapter 5 under the American rule. Everything else equal, binding arbitration will be preferred to trial if (aπ + aδ ) < (tπ + tδ ). Therefore it can be easily demonstrated that the results reached by the model of Chapter 5 under the American rule can be extended to the use of the English rule.

229

. APPENDIX 2

230

References Arianna Andreangeli. Courage Ltd v Crehan and the Enforcement of Article 81 EC before National Courts. European Competition Law Review, 25(12):758, 2004. 11 Homayoon Arfazadeh. Arbitrability under the New York convention: the lex fori revisited. Arbitration International, 17(1):73–87, 2001. 57 Mohammad Reza Baniassadi. Do mandatory rules of public law limit choice of law in international commercial arbitration. Int’l Tax Bus. Law., 10(1):59, 1992. J¨ urgen Basedow. Private enforcement of EC competition law, volume 25. Kluwer law international, 2007. 68 S. Bastianon. L’ arbitrabilit`a delle controversie antitrust tra diritto nazionale e diritto comunitario. Foro it, 1999. 49 L.A. Bebchuk. Litigation and settlement under imperfect information. The RAND Journal of Economics, 15(3):404–415, 1984. URL http://www.jstor. org/stable/10.2307/2555448. Bruce L. Benson. Arbitration. In Encyclopedia of Law and Econoimcs, pages 159–193. 1999. vii P. Bernardini. Stati Uniti: arbitrabilit´a di controversie in materia di legislazione antitrust. Rassegna dell’arbitrato, 3:109, 1985. 46

231

REFERENCES

B. Douglas Bernheim and Michael D. Whinston. Common marketing agency as a device for facilitating collusion. The Rand Journal of Economics, pages 269–281, 1985. 104 Paolo Biavati. Pregiudizale comunitaria e arbitrato. Rivista dell’arbitrato, page 421, 1995. 158, 160 N. Blackaby, C. Partasides, A. Redfern, and M. Hunter. Redfern and Hunter on international arbitration. Oxford University Press, 2009. ISBN 9780199557196. URL http://books.google.be/books?id=AN1EAQAAIAAJ. 1 Gordon Blanke. EC competition law claims in international arbitration. Austrian Arbitration Yearbook 2009, pages 3–92, 2009. 183 Gordon Blanke and Phillip Landolt. EU and US Antitrust Arbitration. A Handbook for Practitioners. Kluwer Law International, September 2010. ISBN 9789041127600. URL http://books.google.nl/books?id=jUzg71Licf4C. 47, 61, 64, 65, 67, 136, 152, 153, 189, 214 Gordon Blanke and Renato Nazzini. Arbitration and ADR of global antitrust disputes: Taking stock (part I). Global Competition Litigation Review, 1(1): 46, 2008a. 33, 41, 49, 51, 83, 191, 192 Gordon Blanke and Renato Nazzini. Arbitration and ADR of global antitrust disputes: Taking stock (part III). Global Competition Litigation Review, 1(3): 133, 2008b. Gordon Blanke and Renato Nazzini. Arbitration and ADR of global antitrust disputes: Taking stock (part II). Global Competition Litigation Review, 1(2): 78, 2008c. Marc Blessing. Mandatory rules of law versus party autonomy in international arbitration. Journal of International Arbitration, 14(4):23–40, December 1997. ISSN 0255-8106. URL https://www.kluwerlawonline.com/abstract.php?area= Journals&id=JOIA1997030. 191

232

REFERENCES

Marc Blessing. Impact of the Extraterritorial Application of Mandatory Rules of Law on International Contracts. Helbing & Lichtenhahn, 1999a. ISBN 9783719018009. Marc Blessing. Introduction to Arbitration: Swiss and International Perspectives. Helbing und Lichtenhahn, 1999b. ISBN 9783719018382. 191 Patrick Joseph Borchers. The choice-of-law revolution: An empirical study. Washington and Lee Law Review, 49:357, 1992. 77 Bryan Caplan and Edward P. Stringham. Privatizing the adjudication of disputes. Theoretical Inquiries in Law, 9(2)(8):1–23, 2008. Dennis W. Carlton and Michael Waldman. The strategic use of tying to preserve and create market power in evolving industries. Technical report, National Bureau of Economic Research, 1998. 104 European Commission. Competition Policy in Europe: The Competition Rules for Supply and Distribution Agreements. EUi collection. Office for Official Publications of the European Communities, 2002. ISBN 9789289439053. URL http://books.google.be/books?id=YYuRAAAAMAAJ. 4 European Commission. Green paper - damages actions for breach of the EC antitrust rules SEC(2005) 1732, December 2005. URL http://eur-lex.europa. eu/LexUriServ/LexUriServ.do?uri=CELEX:52005DC0672:EN:NOT. iv European Commission. White paper on damages actions for breach of the EC antitrust rules SEC(2008) 404 SEC(2008) 405 SEC(2008) 406, April 2008. URL http://eur-lex.europa.eu/LexUriServ/LexUriServ.do?uri=CELEX: 52008DC0165:EN:NOT. iv European Commission. Commission work programme 2012, November 2011. URL http://ec.europa.eu/atwork/pdf/cwp2012 en.pdf. iv James C. Cooper, Luke M. Froeb, Daniel P. O’Brien, and Michael G. Vita. Comparative study of United States and European Union approaches to vertical policy. George Mason Law Review, 13:289,

233

REFERENCES

2004. URL http://heinonline.org/HOL/Page?handle=hein.journals/gmlr13& id=303&div=&collection=journals. 73 James C. Cooper, Luke M. Froeb, Dan O’Brien, and Michael G. Vita. Vertical antitrust policy as a problem of inference. International Journal of Industrial Organization, 23(7):639–664, 2005. 103 Paul Craig and Gr´ainne De B´ urca. EU law: text, cases, and materials. Oxford University Press, 2011. 14, 15, 156 William Laurence Craig, William W. Park, and Jan Paulsson. International Chamber of Commerce arbitration. Oceana Publications, 2000. ISBN 9789284212514. Giuseppe Dari-Mattiacci and Francesco Parisi. Rents, dissipation and lost treasures: Rethinking tullock’s paradox. Public Choice, 124(3):411–422, 2005. URL http://www.springerlink.com/index/YQ83X23576712571.pdf. 155, 164, 171, 173, 177 Dmitry Davydenko and Eugenia Kurzynsky-Singer. Substantive ordre public in Russian case law on the recognition, enforcement and setting aside of international arbitral awards. American Review of International Arbitration, 20(2): 209–233, 2009. 92 Yves Derains. Public policy and the law applicable to the dispute in international arbitration. ICCA Congress Series, No.3:227, 1987. 161 Christopher R. Drahozal and Quentin R. Wittrock. Is there a flight from arbitration? Forum American Bar Association, 6(May):1–46, 2008. vii Thomas Eilmansberger. The relationship between rights and remedies in ec law: in search of the missing link. Common market law review, 41(5):1199–1246, 2004. 12, 15 Theodore Eisenberg and Geoffrey P. Miller. Flight from arbitration: An empirical study of ex ante arbitration clauses in the contracts of publicly held companies, the. DePaul Law Review, 56:335, 2006. URL http://heinonline.org/HOL/Page? handle=hein.journals/deplr56&id=345&div=&collection=journals. vii

234

REFERENCES

J. Faull and A. Nikpay. Faull and Nikpay: The EU Law of Competition. OUP Oxford, 2014. ISBN 9780199665099. URL http://books.google.be/books?id= 3vRpmwEACAAJ. 1 Albert A. Foer. Introduction to symposium on buyer power and antitrust. Antitrust LJ, 72:505, 2004. 101, 103 Vincy Fon and Francesco Parisi. Litigation and the evolution of legal remedies: a dynamic model. Public Choice, 116(3):419–433, 2003. URL http://www. springerlink.com/index/Q20G46141M846024.pdf. Association for International Arbitration. Arbitration in CIS Countries: Current Issues. Maklu, 2012. 95 Jeffry A. Frieden and Paul Kennedy. Global capitalism: Its fall and rise in the twentieth century. WW Norton New York, 2006. 45 Hans Gilliams. Modernisation: from policy to practice. European law review, (4):451–474, 2003. 151 Alexander Greenawalt. Does international arbitration need a mandatory rules method? American Review of International Arbitration, 18:103, 2007. URL http://papers.ssrn.com/abstract=1500663. Warren S. Grimes. Buyer power and retail gatekeeper power: protecting competition and the atomistic seller. Antitrust LJ, 72:563, 2004. 101, 103 Alexander Gurkov. The second-look doctrine in Russian arbitrazh courts. Review of Central and East European Law, 38(3-4):389–402, 2013. 55, 95 Bernard Hanotiau. What law governs the issue of arbitrability? International, 12(4):391–403, 1996. 57

Arbitration

Bernard Hanotiau. International arbitration in a global economy: The challenges of the future. Journal of International Arbitration, 28(2):89–103, 2011. 69 Oliver Hart, Jean Tirole, Dennis W. Carlton, and Oliver E. Williamson. Vertical integration and market foreclosure. Brookings papers on economic activity. Microeconomics, pages 205–286, 1990. 103

235

REFERENCES

Trevor Hartley. The Foundations of European Community Law. Oxford University Press, October 2007. ISBN 9780199290352. 156 Justine S. Hastings and Richard J. Gilbert. Market power, vertical integration and the wholesale price of gasoline*. The Journal of Industrial Economics, 53 (4):469–492, 2005. 104 Daniel Hochstrasser. Choice of law and ”Foreign” mandatory rules in international arbitration. International Arbitration, 11(1):57, 1994. 85 Michael F. Hoellering. The uncitral model law on international commercial arbitration. The International Lawyer, pages 327–339, 1986. 26 Bengt Holmstrom. Moral hazard in teams. The Bell Journal of Economics, pages 324–340, 1982. 102 Howard A. Hood. Extraterritorial application of united states antitrust laws: A selective bibliography, the. Vanderbilt Journal of Transnational Law, 15:765, 1982. URL http://heinonline.org/HOL/Page?handle=hein.journals/vantl15& id=779&div=&collection=journals. 74 Marc Ivaldi, Bruno Jullien, Patrick Rey, Paul Seabright, and Jean Tirole. The economics of unilateral effects. IDEI Working Paper, 222, 2003. 6 Thomas E. Johnson. International antitrust litigation and arbitration clauses. JL & Com., 3:91, 1983. 43 A. Jones and B. Sufrin. EU Competition Law: Text, Cases & Materials. OUP Oxford, 2011. ISBN 9780199572731. URL http://books.google.be/books?id= 0m6cAQAAQBAJ. 1 Clifford A. Jones. Private Enforcement of Antitrust Law in the EU, UK, and USA, volume 3. Oxford University Press New York, 1999. 7 Bruno Jullien and Patrick Rey. Resale price maintenance and collusion. The RAND Journal of Economics, 38(4):983–1001, 2007. 104

236

REFERENCES

Benjamin Klein and Kevin M. Murphy. Vertical restraints as contract enforcement mechanisms. Journal of Law and Economics, pages 265–297, 1988. 102 Assimakis Komninos. Arbitration and the modernisation of European competition law enforcement. World competition, 24(2):211–238, 2001. 152 Assimakis Komninos. Arbitration and EU competition law. SSRN Scholarly Paper ID 1520105, Social Science Research Network, Rochester, NY, April 2009. URL http://papers.ssrn.com/abstract=1520105. 189, 191, 192 V. Korah. An Introductory Guide to EC Competition Law and Practice. Hart, 2004. ISBN 9781841133973. URL http://books.google.be/books?id= hTeFUAwiWfkC. 1 V. Korah and D. O’Sullivan. Distribution Agreements Under the EC Competition Rules. Oxford European Community law series. Hart, 2002. ISBN 9781841132396. URL http://books.google.be/books?id=W8qx5J9lyFwC. 1, 4 Lewis A. Kornhauser. Appeal and supreme courts. In Encyclopedia of Law and Econoimcs, number 1985, pages 45–62. 1999. M. Landes and Richard A. Posner. Adjudication as a private good. Journal of Legal Studies, 8(2):235–284, 1979. Niccol Landi and Catherine A. Rogers. Arbitration of antitrust claims in the united states and europe. SSRN eLibrary, 2007. URL http://papers.ssrn.com/ sol3/papers.cfm?abstract id=962334. 34, 77, 87, 89, 90 Phillip Landolt. Basis for application of competition law in arbitration proceedings. In Practical Aspects of Arbitrating EC Competition Law, T. Zuberbhler and Ch. Oetiker, eds., page 1. Schulthess Juristischee Medien AG, 2007a. 75, 76, 79, 82 Phillip Landolt. Limits on court review of international arbitration awards assessed in light of states’ interests and in particular in light of EU law requirements. Arbitration International, 23(1):63–92, January 2007b. ISSN 0957-

237

REFERENCES

0411. URL https://www.kluwerlawonline.com/abstract.php?area=Journals& id=ARBI2007003. 189 Phillip Louis Landolt. Modernised EC competition law in international arbitration. Kluwer Law International, 2006. 51, 62, 68, 191 White & Case LLP launches in Vancouver. 2010 International Arbitration Survey: Choices in International Arbitration - sponsored by White & Case LLP launches in Vancouver. URL http://www.whitecase.com/articles-10062010/#. Ut0KrPg1ikA. 54 Ashurst (law firm). Study on the conditions of claims for damages in case of infringement of EC competition rules. Technical report. URL http://ec.europa. eu/competition/antitrust/actionsdamages/study.html. Christoph Liebscher. Arbitration and EC competition law - the new competition regulation: back to square one. International Arbitration Law Review, 6(1), 2003a. 152 Christoph Liebscher. The healthy award: challenge in international commercial arbitration. Kluwer Law International, 2003b. ISBN 9041120114 9789041120113. 185, 192 Eric Loquin. La r´eforme du droit fran¸cais interne et international de l’arbitrage. Revue Trimestrielle de Droit Commercial et de Droit Economique, (2):255–288, 2011. 71 Paul Lugard. Ec competition law and arbitration: Opposing principles? European Competition Law Review, 19:295–301, 1998. 48 G. Frank Mathewson and Ralph Albert Winter. An economic theory of vertical restraints. The Rand Journal of Economics, pages 27–38, 1984. 102 Pierre Mayer. Mandatory rules of law in international arbitration. Arbitration International, 2(4):274, 1986. 85, 191 Pierre Mayer. Reflections on the international arbitrator’s duty to apply the law the 2000 freshfields lecture. Arbitration International, 17(3):235–247, January

238

REFERENCES

2001. ISSN 0957-0411. URL http://www.kluwerlawonline.com/abstract.php? area=Journals&id=355301. 190, 191 Joseph T. McLaughlin. Arbitrability: Current trends in the united states. Albany Law Review, 59:905, 1995. URL http://heinonline.org/HOL/Page?handle= hein.journals/albany59&id=929&div=&collection=journals. 86 Geoffrey P. Miller. The legal-economic analysis of comparative civil procedure. The American Journal of Comparative Law, 45(4):905–918, October 1997. ISSN 0002-919X. doi: 10.2307/841025. URL http://www.jstor.org/stable/ 841025. viii Michael J. Mustill and Stewart C. Boyd. The law and practice of commercial arbitration in England. Butterworths London, 1989. 41 Barry Nalebuff. Bundling as an entry barrier. The Quarterly Journal of Economics, pages 159–187, 2004. 104 Horacio Na´on. Unified national legal treatment of international commercial arbitration: a continuing challenge. Arbitration Brief, 2(1), January 2012. URL http://digitalcommons.wcl.american.edu/ab/vol2/iss1/2. 80, 87 Roger G. Noll. ” buyer power” and economic policy. Antitrust Law Journal, pages 589–624, 2005. 101, 103 Daniel P. O’Brien. The antitrust treatment of vertical restraints: Beyond the possibility theorems. The Pros and Cons of Vertical Restraints, 7(1):40–101, 2008. 104 Daniel P. O’Brien and Greg Shaffer. Nonlinear supply contracts, exclusive dealing, and equilibrium market foreclosure. Journal of Economics & Management Strategy, 6(4):755–785, 1997. 104 O’Connor O’Hara, A. Erin, et al. Conflict of laws and choice of law. Conflict of Laws and Choice of Law (November 3, 2009). U Illinois Law & Economics Research Paper No. LE09-030, pages 09–34, 2009. 77

239

REFERENCES

Stanley I. Ornstein. Resale price maintenance and cartels. Antitrust Bull., 30: 401, 1985. 104 Francesco Parisi. Rent-seeking through litigation: adversarial and inquisitorial systems compared. International Review of Law and Economics, 22 (2):193–216, 2002. URL http://www.sciencedirect.com/science/article/pii/ S0144818802000893. Francesco Parisi. Positive, normative and functional schools in law and economics. European Journal of Law and Economics, 18(3):259272, 2004. URL http:// www.springerlink.com/index/B22B8F51F0EE7DFF.pdf. vii Francesco Parisi. Seeking rents in the shadow of coase. American Law & Economics Association Annual, 2008. URL http://law.bepress.com/ cgi/viewcontent.cgi?article=2651&context=alea. Francesco Parisi and Barbara Luppi. Quantitative methods in comparative law. SSRN eLibrary, Minnesota Legal Studies(Research Paper No. 12-20), May 2012. URL http://papers.ssrn.com/sol3/papers.cfm?abstract id=2049907. vii L.A.V.S. Pedro, C.A. Ledesma, J.A.E. S´aenz, J.G. Gilsanz, C.H. Su´arez, J. Al´ munia, A.A. Hern´andez, R.B. Alvarez, L.B. Fuster, C.C. Ortega, et al. Private Enforcement of Competition Law. LEX NOVA, S.A.U., 2011. ISBN 9788498983340. URL http://books.google.be/books?id=AdrAFwOTeioC. 1 Richard A. Posner. What do judges and justices maximize? (The same thing everybody else does). Supreme Court Economic Review, 3:1–41, 1993. Jean-Fran¸cois Poudret and S´ebastien Besson. Comparative law of international arbitration. Sweet & Maxwell, 2007. 51 G.L. Priest and B. Klein. The selection of disputes for litigation. The Journal of Legal Studies, 13(1):1–55, 1984. URL http://www.jstor.org/stable/10.2307/ 724341. 155, 165, 169, 170, 174 PWC and Queen Mary University of London. International Arbitration Survey 2013- Corporate choices in International Arbitration. Technical report, 2013. 54

240

REFERENCES

Jean-Baptiste Racine. L’arbitrage commercial international et l’ordre public. PhD thesis, Nice, 1997. 25 Luca Radicati di Brozolo. Arbitrato, diritto della concorrenza, diritto comunitario e regole di procedura nazionali. Rivista dellarbitrato, 1999. 49, 50 Luca Radicati di Brozolo. Antitrust: A paradigm of the relations between mandatory rules and arbitration–a fresh look at the second look. International Arbitration Law Review, 7(1), 2004. 152 Luca Radicati di Brozolo. L’illic´eit´e qui cr`eve les yeux : crit`ere de contrˆole des sentences au regard de l’ordre public international (`a propos de l’arrˆet Thal`es de la Cour d’appel de Paris). Revue de l’Arbitrage, (3):529, 2005. URL http://www.kluwerarbitration.com/document.aspx?id=ipn26803. 29, 78 Luca Radicati di Brozolo. Arbitration in EC merger control: old wine in a new bottle. European Business Law Review, 19(1):7–15, 2008. 61 Luca Radicati di Brozolo. The impact of national law and courts on international commercial arbitration: mythology, physiology, pathology, remedies and trends. Les Cahiers de l’Arbitrage-The Paris International Arbitration Journal, (3):663–692, 2011. 69 E. Rasmusen. Games and Information: An Introduction to Game Theory. Wiley, 2006. ISBN 9781405136662. URL http://books.google.com.ph/books?id= 5XEMuJwnBmUC. 108, 113 Alan Scott Rau. The arbitrator and mandatory rules of law. American Review of International Arbitration, (Spring), 2008. URL AvailableatSSRN:http://ssrn. com/abstract=1078737. David Reiffen and Andrew N. Kleit. Terminal railroad revisited: Foreclosure of an essential facility or simple horizontal monopoly? Journal of Law and Economics, pages 419–438, 1990. 103, 104 Patrick Rey. Price control in vertical relations. The Pros and Cons of Vertical Restraints, 7(1):135–166, 2008. 104

241

REFERENCES

Patrick Rey and Francisco Caballero-Sanz. The policy implications of the economic analysis of vertical restraints. European Commission, DirectorateGeneral for Economic and Financial Affairs, 1996. 4 Patrick Rey and Jean Tirole. A primer on foreclosure. Handbook of industrial organization, 3:2145–2220, 2007. 103 P.M. Roth, V. Rose, A. Macnab, and C.W. Bellamy. Bellamy & Child Materials on European Community Law of Competition. Oxford University Press, 2013. ISBN 9780199667871. URL http://books.google.be/books?id= OXmwmwEACAAJ. 1 Giesela Ruhl. Methods and Approaches in Choice of Law: An Economic Perspective. Berkeley Journal of International Law, 24(3):801, December 2006. URL http://scholarship.law.berkeley.edu/bjil/vol24/iss3/2http://scholarship. law.berkeley.edu/bjil/vol24/iss3/2/. 39 Michael A. Salinger. Vertical mergers and market foreclosure. The Quarterly Journal of Economics, pages 345–356, 1988. 103 Pieter Sanders, T. M. C. Asser Instituut, and International Council for Commercial Arbitration. Comparative arbitration practice and public policy in arbitration. ICCA Congress Series n.3. Kluwer Law and Taxation Publishers, 1987. ISBN 9789065442963. A. Serravalle. Arbitrato internazionale e leggi antitrust: il caso mitsubishi. Rivista di diritto internazionale privato e processuale, 3:597–616, 1986. 45 S. Shavell. Alternative dispute resolution: An economic analysis. J. Legal Stud., (c):1–22, 1995. URL http://heinonlinebackup.com/hol-cgi-bin/get pdf. cgi?handle=hein.journals/legstud24§ion=5. vii, 198, 202, 204, 205, 206 Natalya Shelkoplyas. The Application Of Ec Law In Arbitration Proceedings. Europa Law Publishing, 2003. ISBN 9789076871165. 84, 85 M. Slade. The effects of vertical restraints: an evidence based approach. The pros and cons of vertical restraints, 2008. 105

242

REFERENCES

Robert H. Smit. Separability and competence-competence in international arbitration: ex nihilo nihil fit? or can something indeed come from nothing? Am. Rev. Int’l Arb., 13:19–197, 2002. 70 Michael Solimine. Economic and empirical analysis of choice of law, an. Ga. L. Rev., 24:49, 1989. 77 Joseph J. Spengler. Vertical integration and antitrust policy. The Journal of Political Economy, pages 347–352, 1950. 102 Martyn D. Taylor. International competition law: a new dimension for the WTO? Cambridge University Press, 2006. 2 Lester G. Telser. Why should manufacturers want fair trade? Journal of law and economics, pages 86–105, 1960. 101, 104 Stuart E. Thiel. Choice of law and the home-court advantage: evidence. American Law and Economics Review, 2(2):291–317, 2000. 77 Antonio Tizzano. Arbitrato privato e competenza pregiudiziale della corte comunitaria. Rassegna dell’arbitrato, (3-4):153–172, 1983. 158, 160 M.A. Utton. Market Dominance and Antitrust Policy. Edward Elgar Publishing, Incorporated, 2005. ISBN 9781843767480. URL http://books.google.it/books? id=5c SAgAAQBAJ. 4 Roger Van den Bergh and Peter D. Camesasca. European Competition Law and Economics: A Comparative Perspective: A Comparitive Perspective. Sweet & Maxwell, London, 2 edition edition, August 2006. ISBN 9780421965805. 3, 4, 9 Walter Van Gerven. Of rights, remedies and procedures. Common Market Law Review, 37(3):501–536, 2000. 15, 16 Hans Van Houtte. The application by arbitrators of articles 81 & 82 and their relationship with the European Commission. European Business Law Review, 19 (1):63–75, February 2008. ISSN 0959-6941. URL http://www.kluwerlawonline. com/abstract.php?area=Journals&id=EULR2008007. 191

243

REFERENCES

Nathalie Voser. Mandatory rules of law as a limitation on the law applicable in international commercial arbitration. American Review of International Arbitration, 7(3-4):319–357, 1997. D. Waelbroeck, E. Clark, and Comisi´on Europea. Direcci´on General de Competencia. Study on the Conditions of Claims for Damages in Case of Infringement of EC Competition Rules. Office for Official Publications of the European Communities, 2004. ISBN 9789289481847. URL http://books.google.be/books?id= K876AAAACAAJ. 15 Gregory J. Werden. Unilateral competitive effects of horizontal mergers: Basic concepts and models. Issues in Competition Law and Policy, 2003. URL http: //papers.ssrn.com/sol3/papers.cfm?abstract id=939033. 6 Michael D. Whinston. Tying, foreclosure, and exclusion. Technical report, National Bureau of Economic Research, 1989. 104 R. Whish and D. Bailey. Competition Law. OUP Oxford, 2012. ISBN 9780199586554. URL http://books.google.be/books?id=9QDhoLiRX-MC. 1 Lawrence J. White. Georgetown study of private antitrust litigation, the. Antitrust Law Journal, 54:59, 1985. URL http://heinonline.org/HOL/Page? handle=hein.journals/antil54&id=157&div=&collection=journals. xii Mark D. White. Theoretical foundations of law and economics. Cambridge University Press, 2009. ISBN 9780521889551. vii Richard Orme Wilberforce, Alan Campbell, Neil Elles, and R. Gresham Cooke. The law of restrictive trade practices and monopolies. Cambridge Univ Press, 1966. 2 Wouter Wils. The optimal enforcement of EC antitrust law: essays in law & economics, volume 33. Kluwer law international, 2002. 13 Wouter Wils. Principles of European antitrust enforcement. Hart Publishing, 2005. 15

244

REFERENCES

Ralph A. Winter. Vertical control and price versus nonprice competition. The Quarterly Journal of Economics, pages 61–76, 1993. 102 Georgios I. Zekos. International commercial and marine arbitration. Routledge, 2008a. 52 Georgios I. Zekos. Antitrust/Competition arbitration in EU versus U.S. law. Journal of International Arbitration, 25(1):1, 2008b. 5, 42, 73 A.N. Zhilsov. Mandatory and public policy rules in international commercial arbitration. Netherlands International Law Review, 42(01):81–119, 1995. doi: 10.1017/S0165070X00003387. Tobias Zuberb¨ uhler and Christian Oetiker. Practical aspects of arbitrating EC competition law. Schulthess, 2007. 153

245